Need and utilization of primary health care among long-term unemployed Finns

Jou nal o Heal h and Social Sciences 2018; 3,3:231-242 231 Need and u iliza ion o p ima y heal h ca e among long- e m unemployed Finns Jan Kla us1, Leena Fo ma2, Jussi Pa anen3, Pekka Rissanen4 ORIGINAL ARTICLE IN HEALTH CARE POLICY KEY WORDS: Deli e y o heal h ca e; heal h ca e u iliza ion; long- e m unemploymen ; p ima y heal h ca e; wo-pa model. Abs ac In oduc ion: Aim o his pape was o iden i y he a ibu es o p ima y heal h ca e u iliza ion among long- e m unemployed Finns, and o examine whe he access o ca e and he choice o p o ide di e wi h espec o employmen s a us. Me hods: Da a on p ima y heal h ca e u iliza ion we e de i ed om wo sou ces; a a ge ed ques ionnai e abou he use o se ices and quali y o li e among long- e m unemployed indi iduals, and he Wel a e and Se ices in Finland Su ey, co e ing he gene al popula ion. A wo-pa econome ic model was applied in o de o sepa a e be ween he p obabili y and le el o u iliza ion. The s a is ical analysis allowed p edic ing he mone a y cos s o p ima y ca e u iliza ion. In his con ex , a non-pa ame ic smea ing ac o was used o adjus o e ans o ma ion bias. In addi ion, a dis inc ion be ween he le el o cos s and numbe o isi s was aken o accoun o he e ec o uni cos a ia ion. Resul s: The analyses indica ed ha he u iliza ion o p ima y heal h ca e se ices among he long- e m unemployed a ied wi h espec o gende , sel - a ed heal h s a us and economic si ua ion, place o esidence, ma i al s a us and du a ion o unemploymen . The scope o analysis was shown o be undamen al o he in e p e a ion o he compa a i e esul s. Taking in o accoun he p o ision o occupa ional ca e se ices in e ed he posi i e e ec o long- e m unemploymen on p ima y ca e u iliza ion. Fu he , while he cos s o u iliza ion we e independen o employmen s a us, long- e m unemploymen had a dis inc educing e ec on he numbe o medical isi s. Discussion and Conclusion: Despi e o g ea e heal h ca e needs, he long- e m unemployed sough less isi s o mo e cos ly public p ima y ca e se ices. In o de o con on unme heal h ca e needs among he long- e m unemployed, public sec o in e en ions should be a ge ed acco dingly, and in pa icula , in ol e gende speci ic social ma ke ing measu es. A ilia ions: 1PhD, Depa men o Heal h Sciences, Facul y o Social Science, Uni e si y o Tampe e, Tampe e, Finland 2PhD, Depa men o Heal h Sciences, Facul y o Social Science, Uni e si y o Tampe e, Tampe e, Finland 3MSc, Depa men o Heal h Sciences, Facul y o Social Science, Uni e si y o Tampe e, Tampe e, Finland 4P o , Depa men o Heal h Sciences, Facul y o Social Science, Uni e si y o Tampe e, Tampe e, Finland Co esponding au ho : D . Jan Kla us, Depa men o Heal h Sciences, Facul y o Social Sciences, Uni e si y o Tampe e, Tampe e, Finland. E-mail: [email p o ec ed] Jou nal o Heal h and Social Sciences 2018; 3,3:231-242 232 Riassun o In oduzione: L’obie i o di ques o s udio è s a o quello di iden i ica e le ca a e is iche dell’u ilizzazione dell’assis enza sani a ia di base a i disoccupa i inlandesi di lunga du a a e quello di esamina e se l’accesso alla cu a e la scel a del o ni o e di e i a ispe o allo s a o di occupazione. Me odi: I da i sull’u ilizzazione dell’assis enza sani a ia di base sono s a i o enu i da due on i: un ques io- na io mi a o sull’uso dei se izi e la quali à della i a a i disoccupa i di lunga du a a e lo S udio Wel a e and Se ices e e ua o in Finlandia sulla popolazione gene ale. Un duplice modello econome ico è s a o applica o pe sepa a e le p obabili à dai li elli di u ilizzazione. L’analisi s a is ica ha consen i o di p e ede e i cos i mone a i dell’u ilizzazione dei se izi di assis enza p ima ia. In aggiun a, una dis inzione a il li ello dei cos i ed il nume o di isi e è s a o a o pe ene e in conside azione la a iazione del cos o uni a io. In ques o con es o il me odo non pa ame ico smea ing è s a o usa o pe ene con o del bias di i as o mazio- ne. In aggiun a, una dis inzione a il li ello dei cos i ed il nume o di isi e è s a o conside a o pe l’e e o di a iazione del cos o uni a io. Risul a i: Le analisi hanno indica o che l’u ilizzazione dei se izi di assis enza sani a ia di base a i disoc- cupa i di lunga du a a a ia a ispe o al gene e, allo s a o di salu e ed economico pe cepi o, alla esidenza, allo s a o coniugale ed alla du a a della disoccupazione. L’analisi si è dimos a a ondamen ale pe l’in e - p e azione dei isul a i compa a i i. Tene e in conside azione la o ni u a dei se izi di assis enza aziendali ha in e i o l’e e o posi i o della disoccupazione di lunga du a a sull’u ilizzazione dei se izi di assis enza p ima ia. Inol e, men e i cos i di u ilizzo e ano indipenden i dallo s a o di occupazione, la disoccupazione di lunga du a a a e a un e e o dis in o di iduzione del nume o delle isi e mediche. Discussione e Conclusione: Nonos an e le maggio i necessi à di cu a, i disoccupa i di lunga du a a si sono so opos i ad un mino nume o di isi e mediche e a se izi di assis enza p ima ia pubblici più cos osi. Pe con on a e i bisogni di assis enza sani a ia imas i insoddis a i a i disoccupa i di lunga du a a, in e en i nel se o e pubblico do ebbe o esse e oppo unamen e mi a i ed in pa icola e coin olge e misu e di so- cial-ma ke ing basa e sul gene e. Compe ing in e es s - none decla ed. Copy igh © 2018 Jan Kla us e al. Edizioni FS Publishe s This is an open access a icle dis ibu ed unde he C ea i e Commons A ibu ion (CC BY 4.0) License, which pe - mi s un es ic ed use, dis ibu ion, and ep oduc ion in any medium, p o ided he o iginal wo k is p ope ly ci ed. See h p:www.c ea i ecommons.o g/licenses/by/4.0/. Ci e his a icle as: Kla us J, Fo ma L, Pa anen J, Rissanen P. Need and u iliza ion o p ima y heal h ca e among long- e m unemployed Finns. J Heal h Soc Sci. 2018;3(3):231-242 TAKE-HOME MESSAGE In Finland, he long- e m unemployed we e ound o ha e unme p ima y heal h ca e needs. In compa ison o he employed popula ion wi h access o occupa ional p ima y ca e, he long- e m unemployed mainly used public p ima y ca e se ices wi h highe ou -o -pocke paymen s and public cos s o p o ision. Recei ed: 18/08/2018 Accep ed: 04/09/2018 Published online: 25/09/2018 DOI 10.19204/2018/ndnd1 Jou nal o Heal h and Social Sciences 2018; 3,3:231-242 233 INTRODUCTION O e he pas h ee decades, popula ion he- al h in Finland has de eloped a o ably. The sha e o he popula ion in good sel -assessed heal h has s eadily inc eased and he inci- dence o long- e m illness has aken a alling end [1–4]. A he same ime, he heal h gap be ween socio-economic g oups has g own wide [5]. The mos ad e se heal h sho ages ha e accumula ed among ma ginal popu- la ion g oups wi h se e al disad an ageous condi ions ela ed o psychological, social and economic esou ces. The unde lying mecha- nisms leading o b oadening heal h gaps a e no s aigh o wa d o iden i y, bu e iden - ly indi iduals being ela i ely disad an aged in e ms o , say income, o en also end o be disad an aged in o he espec s, such as em- ploymen s a us, housing condi ions, social ne wo ks and educa ion. While poo heal h canno solely be a ibu ed o he lack o ac- cess o heal h ca e se ices, a clea di e gence in he choice o se ice p o ide by socioeco- nomic g oups exis s. Acco ding o [6], he use o public p ima y ca e se ices in Finland was s ongly concen a ed in o he lowe income g oups, while he use o occupa ional ca e and p i a e sec o se ices was posi i ely ela ed o income. The numbe o isi s o an occupa- ional ca e gene al p ac i ione was h ee- old in he highes income g oup in compa ison o he lowes income g oup. In addi ion, in- di iduals wi h lowe incomes aced subs an- ially longe wai ing imes o p ima y ca e se ices. A majo pa o hose who had wai- ed un easonably long o access o a medi- cal doc o had wai ed o an appoin men o a public sec o gene al p ac i ione , whe eas appoin men s o occupa ional o p i a e ca e we e usually due wi hin he nex wo days [6]. In Finland, occupa ional heal h ca e se ices a e la gely a ailable o employees a he pu- blic and p i a e sec o s. The s a u o y occupa- ional heal h ca e co e s a wide ange o p e- en i e measu es and in addi ion employe s may o ganize olun a y medical ca e, which is p o ided ee o cha ge o he employees. Abou one-hal o he cos s o occupa ional heal h ca e is inanced by he employe s, he o he hal being inanced by he Social Insu- ance Ins i u ion. In con as o complimen- a y occupa ional ca e heal h se ices, public p ima y ca e se ices o en in ol e a use ee. The maximum ees cha ged o municipal heal h se ices a e de e mined by legisla ion. Municipali ies may op o use lowe a es o p o ide he se ice ee o cha ge. In 2017 he maximum use ee o a public sec o gene- al p ac i ione isi was EUR 20,60 o h ee i s isi s and he ea e addi ional isi s we e complimen a y. Visi s o a public sec o nu - se we e ee o cha ge, and some municipali- ies ( o example Helsinki) p o ided gene al p ac i ione se ices ee o cha ge. Unemploymen inc eased d as ically in Eu o- pean coun ies om he beginning o he e- cession in 2008. Nea ly hal o he 22 million unemployed in he Eu opean Union in 2015 had been ou o wo k o 12 mon hs o lon- ge . As om 2015, he numbe o long- e m unemployed ell by 11% on EU a e age, bu a ia ion ac oss Eu opean coun ies emai- ned subs an ial. While long- e m unemploy- men dec eased in some coun ies (Es onia, Bulga ia, I eland, Poland, UK), i inc eased in F ance, he Ne he lands, Sweden, C oa- ia, Aus ia, La ia, Romania, Luxembou g and Finland [7], 2016). In Finland, long- e m unemploymen began o decline in 2017. In he beginning o 2018, he numbe o long- e m unemployed who had been unemployed o mo e han a yea amoun ed o 80,600, which was 28,600 less han in he p e ious yea . The numbe o long- e m unemployed men dec eased by 16,300 (26%) and ha o women by 12,300 (27%) [8]. In s udies on popula ion heal h, unemploy- men is o en associa ed wi h ad e se men al and physical heal h condi ions [9–12]. Whi- le he unemployed end o be in poo e he- al h han he popula ion in gene al, i would be nai e o conclude ha he ad e se heal h ou comes a ise me ely om a causal ela ion- ship be ween unemploymen and economic ha dship. Ha ing low incomes may a ec physical and men al heal h nega i ely, bu he unde lying mechanisms h ough whi- ch unemploymen may lead o poo e he- Jou nal o Heal h and Social Sciences 2018; 3,3:231-242 234 al h ou comes is known o be a mo e com- plex en i e y [12]. Ill heal h may also cause unemploymen , and hose who a e ela i ely disad an aged in e ms o unemploymen hi- s o ies also end o be disad an aged in o he espec s, such as ha ing low incomes when in wo k, li ing in poo housing condi ions, ha ing low educa ion, pu suing unheal hy li es yles and su e ing om social isola ion [13]. In addi ion, any disce nible ela ionship be ween unemploymen and heal h occu ing a he le el o he popula ion o popula ion sub-g oups is likely o display a high deg ee o a ia ion when examined a mo e disag- g ega ed le els. The pu pose o he p esen pape is o iden- i y he a ibu es o p ima y heal h ca e u i- liza ion among long- e m unemployed Finns, and o examine whe he access o ca e and he choice o p o ide di e wi h espec o employmen s a us. METHODS The da a we e de i ed om he esponses o a long- e m unemploymen ques ionnai e conduc ed in i e la ge- o middle-sized ci ies in Finland (Helsinki; Kuopio; Joensuu; Jy- äskylä; Lappeen an a). A pos ques ionnai e was sen o 1,571 andomly chosen long- e m unemployed esiden s in hese ci ies be ween July and Sep embe 2017. The ques ionnai e esponse a e was 32.5 %, which co esponded o 512 e u ned o ms. In o al 86 cases we e omi ed on he g ounds o ecen e i emen , o he labo ma ke exclusion, o incomple e in o ma ion. The inal da ase consis ed o 426 wo king-aged (21 -65) indi iduals who had been con inuously unemployed o mo e han 12 mon hs p io o he s udy. In o ma- ion on he du a ion o unemploymen was ecei ed om he egis e o Employmen Se ices o Finland (URA-da abase). In he analyses in ol ing bo h he long- e m unemployed (LTU) and he employed popu- la ion, LTU da a was used join ly wi h da a om he Wel a e and Se ices in Finland 2009 su ey (HYPA). The HYPA-su ey was based on a andom sample o 5,800 indi i- duals aged 18–79 yea s. The inal sample size was 3,993, co esponding o a esponse a e o 80%. The combined analysis was a ge ed o hose p ima y ca e se ices and explana- o y a iables ha we e included and asked in he same manne in bo h su eys. Consi- s en ly o he LTU-su ey, only esponden s be ween he age o 21 and 65 we e included in he HYPA da a. In addi ion, s uden s and disabili y pensione s we e excluded, lea ing a o al o 2,843 obse a ions in he inal da a- se . Rele an desc ip i e s a is ics o he wo da ase s a e p esen ed in Table 1. The analysis ocused on he u iliza ion o he ollowing p ima y ca e se ices: 1) public ge- ne al p ac i ione ; 2) public heal h nu se; 3) occupa ional ca e gene al p ac i ione ; and 4) occupa ional ca e heal h nu se. Occupa io- nal heal h ca e was ega ded a subs i u e o public p ima y ca e, as employees co e ed by occupa ional ca e ha e li le o no incen i es o using public p ima y ca e se ices. In o - de o accoun o he subs i u ion e ec , he analysis on he in e ela ed da a was ca ied ou i s , o public p ima y ca e se ices only and secondly, o p ima y ca e se ices inclu- ding occupa ional heal h ca e. On g ounds o una ailabili y o in o ma ion on he special y o p i a e sec o physician isi s, hey we e no included in he analysis. Heal h ca e u iliza ion da a ha e a subs an ial p opo ion o alues a ze o, and hence he econome ic modeling o such “ze o-in la ed” da a is no s aigh o wa d. Se e al ela ed es ima ion me hods ha e been p oposed, in- cluding he Tobi model, he wo-pa model and he sample selec ion model [14]. In he p esen s udy, he wo-pa model was consi- de ed app op ia e. In he i s pa , a logi mo- del o bina y esponses was i ed on he en- i e da a, gi ing an es ima e o he p obabili y ha an indi idual possessing one o se e al o he cha ac e is ics speci ied in he model had used p ima y heal h ca e se ices. The second pa applied OLS eg ession o es ima e he e ec o he explana o y a iables on he le el o use o indi iduals wi h non-ze o u iliza- ion. The expec ed le el o p ima y heal h ca e use o an indi idual possessing pa icula model inclusi e cha ac e is ics was calcula ed Jou nal o Heal h and Social Sciences 2018; 3,3:231-242 235 o example cases. The non-ze o obse a ions a e usually no no mally dis ibu ed as hey end o be he- a ily skewed o he igh . The e o e, a log ans o ma ion o he dependen a iable is commonly unde aken in heal h ca e u ili- za ion analysis. Besides possessing desi able s a is ical p ope ies, he semi-log o dou- ble-log speci ica ions gene a e es ima es wi h s aigh o wa d economic in e p e a ions. In addi ion, a log ans o ma ion sho ens he long igh ail, lessens he e oscedas ici y, and dec eases he in luence o ou lie s. The un- c ional o m o he models o p ima y ca e u iliza ion was es ed in he Box-Cox a- mewo k. In all models he loga i hmic an- s o ma ion o he dependen a iable was s a- is ically suppo ed (i.e. he null hypo hesis ha he models a e he same in e ms o go- odness o i was ejec ed by he c i ical alue o Chi-squa ed a 5 % le el in a o o he semi-log model). As in addi ion o es ima ing he e ec o he explana o y a iables on heal h ca e u iliza- ion, he pu pose o he s udy was o p edi- c he le el o u iliza ion, he es ima es on a log-scale we e e ans o med o he o iginal scale. A non-pa ame ic smea ing ac o was used o ake in o accoun o e ans o ma- ion bias [15, 16]. The dependen a iables in he logi mo- dels we e he use/non-use o public gene al p ac i ione , public heal h nu se, occupa ional ca e gene al p ac i ione and occupa ional ca e heal h nu se se ices. The subs i u ion models we e es ima ed o a dependen a- iable whe e he u iliza ion o occupa ional ca e se ices was ega ded as a subs i u e o public p ima y ca e u iliza ion o indi iduals in employmen . In he le el models, he mo- ne a y alue o he dependen a iables was calcula ed o indi iduals wi h non-ze o u i- liza ion by mul iplying he o al numbe o isi s by he uni cos o he se ices [17]. As he uni cos s o occupa ional ca e se ices we e lowe han hose o public p ima y ca e, he espec i e le el models we e also es ima- ed o he numbe o isi s. In he o he le el models he le el o cos s and he le el o isi s we e in e changeable as a cons an uni cos was used o his u iliza ion da a. The independen a iables included in he LTU p obabili y models we e dummies o : gende ( emale), age (35-54; 55-65), egion (Helsinki), sel - a ed heal h s a us (bad/ e y bad) and economic si ua ion (di icul / e y di icul ). As o he sel - a ed heal h a iable, a i e-i em scale was included in he que- s ionnai e (good, a he good, a e age, bad o e y bad). The ques ion abou household economic si ua ion conce ned he easiness o co e ing compulso y household expendi u es ( e y easy, easy, a he easy, a he di icul , di - icul o e y di icul ). The same se o expla- na o y a iables was applied in he le el mo- dels, excep o sel - a ed heal h s a us, which u ned ou o be an insigni ican p edic o o he le el o p ima y ca e u iliza ion. In addi- ion, a dummy o ma i al s a us (ma ied) and a con inuous a iable o unemploymen du a ion (UNEMPDUR) we e included. Fo he p obabili y models es ima ed on he combined (LTU and HYPA) da a, he inde- penden a iables we e: gende ( emale), age (35-54; 55-65), sel - a ed heal h s a us (bad/ e y bad), economic si ua ion (di icul / e y di icul ) and a dummy o long- e m unem- ploymen (LTU). In speci ica ions o he le el models, an iden ical se o explana o y a iables was suppo ed and used in he es i- ma ions. RESULTS Pa e ns o p ima y ca e u iliza ion A g aphical illus a ion o non-ze o u iliza- ion o ou pa ien se ices among he long- e m unemployed (LTU) and he gene al po- pula ion in employmen (HYPA) is p esen ed in Figu e 1. This se ing e lec s he baseline o he p obabili y models in he o hcoming econome ic analyses. Mo e han 60% o he LTU had used public gene al p ac i ione se - ices in he p eceding 12 mon hs, while he co esponding sha e o hose in employmen was abou 30%. The LTU also had a subs an- ially highe pe cen age o public heal h nu se se ices use (45% s. 23%), and a sligh ly hi- Jou nal o Heal h and Social Sciences 2018; 3,3:231-242 236 ghe use o ou pa ien clinic specialis se i- ces (31% s. 23%). One- hi d o he employed had used occupa ional ca e gene al p ac i io- ne se ices and abou one- ou h occupa io- nal heal h nu se se ices. As indi iduals wi h pe manen employmen we e likely o ha e subs i u ed hese se ices o public p ima y ca e se ices, he o al pe cen age o gene al p ac i ione (65%) and heal h nu se se ices (50%) use u ned ou sligh ly highe o he employed. As ega ds he le el o use, hose employed who had used public gene al p ac i ione se ices, had made mo e isi s o hem han he LTU (1.9 s 1.6) (Figu e 2). Assuming ha occupa ional p ima y heal h ca e se i- ces we e subs i u es o public p ima y ca e o he employed, he di e ence in he o al numbe o isi s o a gene al p ac i ione in- c eased u he (2.8 s 1.6). The numbe o isi s o a public heal h nu se was equal (1.8), whils accoun ing o subs i u ion inc eased he o al numbe o heal h nu se isi s o he employed om 1.8 o 2.3. The employed had also consul ed mo e o en ou pa ien clinic specialis s and p i a e doc o s. Econome ic analysis o p ima y ca e u ili- za ion Fo he long- e m unemployed, heal h s a us, as measu ed by sel -assessed heal h, had a di- s inc e ec on he p obabili y o p ima y ca e use (Table 2). As indica ed by he logi coe - icien s, poo heal h was posi i ely ela ed o he p obabili y o ha ing sough public sec o ca e om a gene al p ac i ione o a heal h nu se. In compa ison o a e e ence indi idual in good heal h, an indi idual in poo heal h was abou 1.5 imes mo e likely o ha e con- sul ed a gene al p ac i ione o a heal h nu - se. Ano he signi ican explana o y a iable o simila magni ude was di icul economic si ua ion, which inc eased he p obabili y o p ima y ca e use in all models excep heal- h nu se se ices, whe e he ela ionship was s ill posi i e, bu no s a is ically suppo ed. Female gende was posi i ely ela ed o ha- ing used gene al p ac i ione se ices, while no e ec was ound o heal h nu se se ices. A emale wi h o he wise simila cha ac e i- Table 1. Desc ip i e s a is ics o LTU and HYPA da a. LTU HYPA N426 2843 Age (mean) 51.0 45.6 Gende (%) Male 53.9 47.6 Female 46.1 52.4 Sel -assessed heal h (%) LTU Male Female To al Good 5.3 5.7 5.4 Ra he good 42.3 39.7 41.0 A e age 19.8 24.7 22.2 Bad 24.7 22.7 23.6 Ve y bad 7.9 7.2 7.8 HYPA Male Female To al Good 42.7 45.2 43.9 Ra he good 33.6 33.1 33.3 A e age 19.0 17.9 18.5 Bad 3.9 2.9 3.4 Ve y bad 0.8 0.9 0.9 Jou nal o Heal h and Social Sciences 2018; 3,3:231-242 237 Figu e 1. Pe cen age o indi iduals wi h ou pa ien heal h se ices use in las 12 mon hs (LTU = long- e m unem- ployed; HYPA-2009 = gene al popula ion in employmen ). Figu e 2. Numbe o isi s o ou pa ien heal h se ices among he long- e m unemployed (LTU) and he gene al popula ion in employmen (HYPA-2009). s ics as he e e ence male, had a 1.8 imes highe p obabili y o gene al p ac i ione use. Li ing in he capi al Helsinki had a sligh in- c easing e ec on he p obabili y o seeking gene al p ac i ione ca e. By con as , age had no disce nible in luence on he likelihood o p ima y ca e use. Howe e , as ega ds he le el o use, he long- e m unemployed in he oldes age g oup used less gene al p ac i ione and heal h nu - se se ices in compa ison o he e e ence age g oup. A nega i e e ec on he use/cos s o gene al p ac i ione se ices also applied o he du a ion o unemploymen . The longe unemploymen had las ed he less isi s o a gene al p ac i ione ca e occu ed. While ma- i al s a us was an insigni ican indica o o he p obabili y o use, and was hus, excluded om he logi models, being ma ied had an ob ious inc easing e ec on he le el o p i- ma y ca e use. Es ima ion esul s on he join analysis exclu- ding occupa ional ca e subs i u ion a e p e- sen ed on he le -hand side o Table 3. These models we e es ic ed o include only ho- se employed indi iduals who had due o he Jou nal o Heal h and Social Sciences 2018; 3,3:231-242 238 una ailabili y o occupa ional heal h ca e, o o he easons, used public gene al p ac i ione o heal h nu se se ices. Again, emale gen- de was a s ong p edic o o ha ing sough p ima y ca e, bu had no e ec on he le el o use a e he p ima y con ac . The e ec o olde age was wo- old; in compa ison o he younges age g oup, hose be ween ages 35- 54 we e less likely o ha e used public gene al p ac i ione and heal h nu se se ices, whe- eas an inc easing e ec applied o hose in he age g oup 55-65. This could be due o a be e co e age o occupa ional ca e o indi- iduals wi h well-es ablished occupa ions and ela i ely li le heal h p oblems. As expec ed, poo heal h was s ongly associa ed wi h bo h he p obabili y and le el o p ima y ca e use. Long- e m unemploymen was associa ed wi h a highly ele a ed p obabili y o gene al p ac i ione and heal h nu se use, and while some indica ion o a lesse quan i y o use o hese se ices was implied, he ela ionship was no s a is ically suppo ed. As ega ds he p obabili y and cos s o public p ima y ca e use, a young unemployed emale in poo heal h had a p obabili y o gene al p ac i io- ne use nine imes he p obabili y o he e- e ence indi idual (young employed male in good heal h). The co esponding p obabili y o public heal h nu se use was six- old. The expec ed public sec o cos s o o al p ima y ca e u iliza ion in ela ion o he e e ence indi idual we e EUR 770 and EUR 562, e- spec i ely. Allowing o occupa ional heal h ca e subs i- u ion had a ma ked e ec on he pa e ns o gene al p ac i ione and heal h nu se u iliza- ion (Table 3, igh hand side). Gende and poo heal h s ill s ongly inc eased he p o- babili y o using hese se ices, whe eas in hese models also olde age had posi i e sign in bo h age g oups. The e ec o long- e m unemploymen changed o he opposi e. Being long- e m unemployed had a educing e ec on he p obabili y o gene al p ac i io- ne and heal h nu se use when compa able se ice use aking place in occupa ional he- al h ca e was included in he analysis. Long- e m unemploymen had no disce nible e ec on he le el o cos s o gene al p ac i ione Table 2. Es ima ion esul s o he LTU p obabili y (logi ) and le el models (OLS). Va iable Gene al p ac i ione Heal h nu se P obabili y o u iliza ion Coe . P>z Coe . P>z Gende ( emale) 0.752*** 0.000 0.069 0.731 Age (35-54) 0.316 0.465 0.006 0.989 Age (55-65) 0.468 0.278 -,035 0.933 Region (Helsinki) 0.551* 0.036 -0.359 0.135 Hela h S a us (bad/ e y bad) 0.511* 0.017 0.479* 0.019 Economic si ua ion (di icul / e y di icul ) 0.571** 0.009 0.364 0.079 Cons an -0.838 0.062 -0.569 0.191 N 426 426 Le el o cos s Coe . P> Coe . P> Gende ( emale) -0.007 0.932 -0.058 0.628 Age (35-54) -0.303 0.092 -0.406 0.077 Age (55-65) -0.365* 0.042 -0.697** 0.003 Economic si ua ion (di icul / e y di icul ) 0.161 0.058 0.206 0.087 Unempdu -0.094* 0.036 0.021 0.757 Ma i al s a us (ma ied) 0.265** 0.005 0.420** 0.003 Cons an 6.093*** 0.000 4.728*** 0.000 N 262 163 *** P < ,001; ** P < ,01; * P < .05. Jou nal o Heal h and Social Sciences 2018; 3,3:231-242 239 Table 3. Es ima ion esul s o he p obabili y (logi ) and le el models (OLS). Va iable Gene al p ac i ione (public sec o ) Gene al p ac i ione (inc. occupa ional ca e) P obabili y o u iliza ion Coe . P>z Coe . P>z Gende ( emale) 0.510*** 0.000 0.504*** 0.000 Age (35-54) -0.216* 0.030 0.252** 0.005 Age (55-65) 0.290** 0.006 0.356*** 0.000 LTU 0.815*** 0.000 -0.105 0.391 Hela h S a us (bad/ e y bad) 0.901*** 0.000 0.750*** 0.000 Economic si ua ion (di icul / e y di icul ) 0.299** 0.013 0.039 0.738 Cons an -1.148*** 0.000 -0.300*** 0.000 N 3269 3269 Le el o cos s Coe . P> Coe . P> Gende ( emale) 0.043 0.281 0.132*** 0.000 Age (35-54) -0.038 0.480 -0.128** 0.006 Age (55-65) -0.063 0.241 -0.049 0.328 LTU -0.105 0.056 0.058 0.327 Hela h S a us (bad/ e y bad) 0.323*** 0.000 0.439*** 0.000 Economic si ua ion (di icul / e y di icul ) 0.097 0.077 0.150** 0.007 Cons an 5.366*** 0.000 5.102*** 0.000 N 1157 1833 Heal h nu se (public sec o ) Heal h nu se (inc. occupa ional ca e) P obabili y o u iliza ion Coe . P>z Coe . P>z Gende ( emale) 0.431*** 0.000 0.313*** 0.000 Age (35-54) -0.658 *** 0.000 -0.052 0.568 Age (55-65) -0.189 0.096 -0.147 0.138 LTU 0.586*** 0.000 -0.399*** 0.001 Hela h S a us (bad/ e y bad) 0.780*** 0.000 0.482*** 0.000 Economic si ua ion (di icul / e y di icul ) 0.214 0.093 -0.78 0.496 Cons an -1.320*** 0.000 -0.286*** 0.000 N 3269 3269 Le el o u iliza ion Coe . P> Coe . P> Gende ( emale) 0.111 0.062 0.127** 0.003 Age (35-54) -0.330*** 0.000 -0.336*** 0.000 Age (55-65) -0.279*** 0.000 -0.162** 0.006 LTU -0.100 0.218 0.164* 0.033 Hela h S a us (bad/ e y bad) 0.416*** 0.000 0.561*** 0.000 Economic si ua ion (di icul / e y di icul ) 0.058 0.463 0.146* 0.038 Cons an 4.627*** 0.000 4.185*** 0.000 N 764 1462 *** P < ,001; ** P < ,01; * P < .05.