Geographic Distribution of Health Care - Regional differences in demand and supply, costs and quality (GeoHealth)

There are significant regional differences in the use of health services within OECD countries, but we know little about the causes and consequences of this. It is important to understand the mechanisms behind the regional differences in order to be able to develop policy measures that can meet the needs of an aging, multicultural and relatively decentralized Norwegian population. We studied regional variation in travel distances, use of health services and patients' health outcomes. Since the travel distance and the consequences of this can be affected by policy measures, perhaps most easily by the geographical location of health services, our analyses are directly relevant to policy makers. Securing qualified health personnel and knowledge of factors that determine where they work geographically is important in meeting the expected increase in the need for health services for the aging population in most OECD countries. Since the opportunities for informal care are also essential for health and care needs, we investigated the significance of distance between elderly people in need of help and their adult children for the use of care services in a regional perspective. Our analyzes combine several large Norwegian register datasets. We used the geographical location of the place of residence of all Norwegian residents and health institutions (hospital, GP, etc.), and calculated the travel distance between the places. We also have data for educational attainment and workplace, and we have several indicators of the use of health services and health outcomes (consultations, hospital admissions, disability benefits, employment, mortality, etc.). Our results show that there is a big difference in the use of health services between public hospitals. Factors linked to hospitals explain around half of this difference, while the rest can be explained by the patients' background. There is also great geographical variation in the use of private hospitals financed by the public sector. Free treatment choice (FBV) was introduced in November 2015 and opened the way for private providers to become approved suppliers of selected treatments, thereby treating patients at the expense of the state. Our analyzes show that FBV is used almost exclusively by patients living near the large cities in the south of Norway. It is reasonable to link this to the strong geographical concentration of where the private providers have chosen to establish themselves. We also find that people who use FBV are more often young and have higher education. However, the explanations for this are more complex. Our results also show great geographical variation in capacity occupancy at maternity wards. We find that women who enter maternity clinics on more hectic days receive fewer and less invasive treatments, and their children get at least as good APGAR scores. We also do studies on the primary healthcare service. We find that the median travel time from home to the GP is just under five minutes, while the travel time of the ninetieth percentile was just over 18 minutes. Even in the sparsely populated areas of the country, travel times were relatively low, with five minutes on the median, and under 45 minutes on the ninetieth percentile. Nevertheless, we find that people move closer to their GP when they start using these services, and particularly among those who live far away. Furthermore, we find that patients use the website legelisten.no when choosing a GP. Patients choose doctors who have a higher rating, and it is particularly women and patients with a high socio-economic status who make use of the online information. We also find that the number of visits to the GP is drastically reduced when children turn 16 and thus have to pay a deductible. The reduction is greatest for children who live close to their GP. This means that those with a short travel distance to service providers are more sensitive to prices in the health service. With regard to the interaction between informal and formal care, we found that elderly without resourceful relatives have a greater need for care services than those with partners and children, but that the distance to adult children was not significant. In a regional perspective, there were only small differences between central and less central municipalities, and also small differences between municipalities with different resource situations, both economically and demographically.

I prosjektet har vi hentet lærdom og inspirasjon fra nasjonale og internasjonale forskere innen helseøkonomi, samt representanter fra brukersiden. Vi avholdt derfor jevnlig workshoper der vi inviterte ledende forskere innen fagfeltet; fra andre fagmiljøer i Norge, samt Norden forøvrig og USA, i tillegg til helsepersonell (særlig sykepleiere og leger) med særlig kunnskap om det norske helsesystemet og dets potensielle utfordringer. Både inviterte forskere samt prosjektets egne forskere presenterte ved disse anledningene eget arbeid, og det ble satt av tid til å diskutere institusjonelle og metodiske utfordringer, samt mulige samarbeidsprosjekter. I tillegg har flere av prosjektdeltakerne tilbrakt tid i utlandet, noe som har bidratt til å danne kontaktnett over landegrensene. Workshopene og utenlandsopphold har således betydd mye for prosjektdeltakernes faglige utvikling i løpet av prosjektperioden, men også vært viktig for å skape et kontaktnett som kan dras nytte av i videre karriere. Også samarbeidet med brukergruppen - helsepersonell på ulike nivåer og byråkrater - har vært nyttig for å gi institusjonell kunnskap om fagfeltet, og har vært en viktig plattform for videre samarbeid. Med hensyn til konsekvenser for helsetjenester på ulike nivåer ne på kort sikt, både nasjonalt og regionalt, er nytten av mer indirekte karakter. For kommunale omsorgstjenester er et viktig funn at omsorgstjenester synes å fordeles rimelig ‘rettferdig’ i Norge, også i et regionalt perspektiv. Flere av artiklene har mottatt stor interesse fra fagfolk fra ulike disipliner og miljøer underveis i prosjektet. Disse har blitt presentert blant annet for arbeidsgrupper på Folkehelseinstituttet, Helsedirektoratet, og Institutt for sykepleie ved OsloMet. Arbeidet med bruk av digitale plattformer i primærhelsetjenesten har blitt brukt i avgjørelser i høyesterett. Prosjektet har også lagt grobunn for samarbeid og utvikling av offisiell statistikk i Statistisk Sentralbyrå. Vi har diskutert mulighet for publisering av ulike nye offisielle statistikker, som reiseavstand til fastlege og akuttjenester, og samarbeidet om bruk av GIS Geografiske data på både personer og ulike helsetjenester, samt mulighetene for å koble individer på tvers av generasjoner og i parforhold muliggjør unike framtidige prosjekter. Gitt den forventede demografiske utviklingen med sterk aldring og en mer heterogen befolkning vil både helsetjenester og muligheter for uformell omsorg bli en knapphetsgode framover, og da særlig på mindre sentrale steder. Sosial og regional ulikhet i helse kan således forventes å øke. Kunnskapen fra dette prosjektet utgjør et godt utgangspunkt for framtidige prosjekter med et enda sterkere fokus på muligheter for politikktiltak. Dette prosjektet har vist at å koble store mengder data er mulig (om enn tidkrevende), og således har prosjektet synliggjort viktigheten av fortsatt mulighet for store og sensitive registerkoblinger på helsefeltet.

Substantial regional differences exist in the utilization of health care within OECD countries, but knowledge remains scarce on the causes and consequences thereof. Understanding mechanisms behind the regional differences is crucial to enable the construction of efficient policies to meet the surging needs of an aging, multicultural and relatively decentralized Norwegian population. We will provide policy relevant knowledge on effects of travel distances on regional distribution of health care personnel, on regional utilization of health care (primary and specialist) and on possible regional variation in indicators of patient recovery. Since travel distance and consequences thereof can be affected by policy, perhaps most easily by the geographic location of health care services, our suggested analyses will be directly relevant for policy design. Indeed, securing qualified personnel, and knowledge on what makes them allocate rurally, may be one of the most important challenges in meeting the expected surge in health care needs of aging OECD-populations. We will combine several population-wide, individual-level data registers to undertake these novel analyses. Using the exact geographic location of the home of all Norwegian residents and health care providers (hospital, office of one's general practitioner, etc.), we calculate travel distance between the locations. We also know everyone's education and where they work, and we have several (admittedly imperfect) indicators of health care utilization and health outcomes, by health care provider (consultations, hospitalizations, disability benefits, employment, mortality, etc.). The extremely rich data allow us to use plausibly exogenous variation and quasi-experimental methods to reliably estimate causal effects. The research team has extensive experience in undertaking top-quality research on these data registers.