Abstraktbok. Nationella hälsoekonomiska konferensen mars Ekonomihögskolan Lund

Transkript

1 Abstraktbok Nationella hälsoekonomiska konferensen mars Ekonomihögskolan Lund Svensk Förening för Hälsoekonomi (SHEA) i samverkan med Ekonomihögskolan vid Lunds universitet

2 INNEHÅLL PROGRAMMET I KORTHET... 5 FÖREDRAG I PLENUM... 6 Economics of aging and elderly care... 6 The future of health economics development and financing: Centers for Health Economic Research in Germany the case of Hamburg... 7 TEMASESSIONER... 8 Temasession A: Research on the Swedish Health Care System... 8 Temasession B: (Re)considering a coherent approach to operationalize an appropriate perspective in economic evaluation in Sweden... 9 Temasession C: Systematiska översikter av hälsoekonomiska utvärderingar Hur gör man och vad är det värt? Temasession D: Validering av hälsoekonomiska modeller varför, hur och för vem? MUNTLIGA FÖREDRAG Parallellsession A Forgetting to remember or remembering to forget -a study of the recall period length in health care survey questions Short- and Long-Term Effects of Free Care Services for Mothers and Infants - Evidence from a Field Experiment Early-life exposure to adverse disease environment and labor force exit: Southern Sweden The impact of a childhood onset of Type 1 diabetes on higher education and labour market outcomes Parallellsession B En utvärdering av det statliga tandvårdsstödet rapport från TLV:s regeringsuppdrag Towards a consistent use of Cost of Added Life-Years by TLV Betydelsen av kostnadseffektivitet och sjukdomssvårighet för subvention av läkemedel i Sverige: en analys av Tandvårds- och läkemedelsförmånsverkets beslutsfattande Hälsoekonomiska analyser för insatser till sköra äldre SBU:s projekt inom äldreområdet Parallellsession C Is the level of patient co-payment for medicines associated with refill adherence in Sweden? Pharmaceuticals An analysis based on Swedish real-life data A new approach to decomposing the concentration index of health using the recentered influence function Income receipt and mortality - evidence from Swedish public sector employees Parallellsession D

3 1. SWEDISH EXPERIENCE-BASED VALUE SETS FOR EQ-5D HEALTH STATES The implications of experience-based versus hypothetical health state valuations in health economic evaluations Changes in lifestyle risk factors: Health and economic impact as estimated by the population based RSH model EVPI CURVES IN PRACTICE Parallellsession E Identifiering av sjukdomsfenotyper, och hur de påverkar interventioners kostnadseffektivitet vid Crohns sjukdom Cost-utility analysis of treatment for acute anterior cruciate ligament tears: a randomized controlled trial Prevention of cannabis among Swedish teens assessing the cost-effectiveness of a schoolbased prevention program using Markov modelling Aktiv hälsostyrning ny modell för riktad prevention av oplanerad sjukvård Parallellsession F Hur har införandet av vårdval påverkat patientnöjdheten i primärvården? Learning-by-Doing in a High-Skill Profession when Stakes are High: Evidence from Advanced Cancer Surgery Costly continuity? An analysis of care continuity and care consumption in Sweden Fooled by the means - opposite distributional trends in GP visits when using individual v.s. area level income data POSTERS Postervandring A LABOR MARKET CONSEQUENCES OF GROWING UP WITH TYPE 1 DIABETES Excess costs of medical care 1, 8, 15 and 24 years after diagnosis of diabetes: estimates from young and middle aged incidence cohorts in Sweden Estimating and exploring health utilities of type 2 diabetes-related complications: a crosssectional study in Sweden Cost-effectiveness analysis of wet and ready-to-use catheters Resource use in patients with psoriasis after the introduction of biologics in Sweden - Direct costs increase, whereas indirect costs decrease Kostnadseffektivitet av allmän vaccination mot rotavirus Costs and health outcomes from surgery in rural Uganda Postervandring B Lean Accounting limits in healthcare - A case study at SUS Hälsoekonomisk medvetenhet bland tandläkarstudenter

4 3. Is it worth the money? A health economic evaluation of national parenting programs in Sweden Hälsoekonomisk utvärdering av intensiv multimodal rehabilitering för patienter med kronisk smärta How Does the Revelation of Ill-Health Affect Individuals' Subjective Well-Being? Evidence from a Screening-Program for Abdominal Aortic Aneurysm Using a Regression Discontinuity Design Patient perspectives on centralisation of low volume, highly specialised procedures in Sweden Introduktion av klinikläkemedel i vår omvärld vilka goda exempel finns att lyfta fram

5 PROGRAMMET I KORTHET Torsdag 20 mars 11:30-12:30 Registrering, lunchsmörgås och kaffe 12:45-14:15 Plenum Ekonomihögskolan hälsar välkommen Economics of aging and elderly care Edward C Norton, University of Michigan Paneldiskussion 14:15-14:45 Kaffe och posterutställning 14:45-16:00 Temasessioner A, B 16:00-16:15 Kaffe och posterutställning 16:15-17:45 Parallellsessioner A-C 17:45 SHEA årsmöte 19:00 Middag Fredag 21 mars 07:45-08:15 Registrering och kaffe 08:15-09:45 Parallellsessioner D-F 09:45-10:15 Kaffe och posterutställning 10:15-11:15 Postersessioner A, B 11:15-12:30 Temasessioner C, D 12:30-13:15 Lunchsmörgås och kaffe 13:15-14:45 Plenum The future of health economics development and financing Centers for health economic research in Germany the case of Hamburg Jonas Schreyögg, Hamburg University, Hamburg Center for Health Economics Paneldiskussion 14:45-15:00 Avslutning 5

6 FÖREDRAG I PLENUM Economics of aging and elderly care Edward C Norton, University of Michigan : Financing long-term care will be a growing problem over the next several decades in developed countries as the fraction of elderly persons increase. Home health care, which is a substitute and a complement to other kinds of health care, provides an important lesson in the challenges of financing long-term care with prospective payment. Prospective payment combats moral hazard in health care for high users, but can create moral hazard problems for low users at both the extensive and intensive margins. The prospective payment system for home health care, created by Centers for Medicare & Medicaid Services to reduce total expenditures, instead provided perverse incentives that predictably increased the number patients, episodes, visits per episode, and total expenditures among low users. We call this effect low-user moral hazard. The effects were particularly pronounced among for-profit agencies. Furthermore, we show that the influence of forprofit market share increased over time in the home health care market, and propose three economic explanations for the increase. These explanations include the entry of new agencies to the market, the influence of new agencies on existing agencies, and the fact that chain-affiliated agencies learned profitable service provision patterns from other agencies in the same chain. Our results have implications for current policies in the Affordable Care Act. Paneldiskussion Moderator: Carl Hampus Lyttkens Diskutander: Erik Grönqvist, IFAU och Uppsala universitet, Katarina Steen Carlsson, Lunds universitet och IHE, Nils Janlöv, Myndigheten för vårdanalys, Thomas Davidson, SBU, Sölve Elmståhl, Lunds universitet Tid: Torsdag 20 mars 12:45-14:15 6

7 The future of health economics development and financing: Centers for Health Economic Research in Germany the case of Hamburg Jonas Schreyögg, Hamburg University, Hamburg Center for Health Economics : Health Economics as a scientific discipline has developed rapidly over the last decades. The development of health economics was mainly driven by Anglo-American countries in the 90ties. Many universities in the US and the UK have established centers for health economics or at least centers for health services research including health economics. During the last 10 years other European countries followed and also developed centers for Health Economics. Experience from these countries shows that research output in health economics will benefit from a concentration of scholars with different health economic research fields and different levels of experience. To date health economic research in Germany is highly fragmented. For this reason the German Ministry of Education and Research (BMBF) has decided to publish a call for funding of centers for health economic research in Germany. Every university or research institution in Germany with intention to establish a center for health economics was welcome to apply for this grant. The institutions were required to reveal their current capacities and future plans for expansion. The overall aim of the funding programme was to provide the institutions with seed money that they are supposed to use for establishing the necessary structures for health economic research centers. The Hamburg Center for Health Economics (HCHE) has been selected as one of these centers. It was established as a joint center of the Medical Faculty and the School of Business, Economics and Social Sciences at the University of Hamburg. The German approach as well as the Hamburg Center for Health Economics (HCHE) will be presented in this session. Paneldiskussion Moderator: Birger Forsberg, Stockholms läns landsting Medverkande: Per Carlsson, Linköpings universitet, Joakim Ramsberg, Myndigheten för vårdanalys, Ulf Persson, IHE, Carl Hampus Lyttkens, Lunds universitet, Gabriella Chirico, Uppsala universitet, Emelie Heintz, SBU, Margareta Dackehag, Lunds universitet. Tid: Fredag 21 mars 13:15-14:45 7

8 TEMASESSIONER Temasession A: Research on the Swedish Health Care System Organisatör: Björn Ekman, Lunds universitet Tid: Torsdag 20 mars 14:45-16:00 Moderator: Björn Ekman, Lunds universitet Medverkande: Anders Anell, Lunds universitet, Clas Rehnberg, Karolinska institutet, Richard Saltman, Emory University och European Observatory on Health Systems and Policies This thematic session brings together the experiences from three health care analysts, Professor Anders Anell, Lund University, Professor Clas Rehnberg, Karolinska Institute, and Professor Richard Saltman, Emory University and European Observatory on Health Systems and Policies. The particular presentations are: 1. Professor Anell: Effekter av vårdval och patientinflytande 2. Professor Rehnberg: The EuroHOPE research project on European Health Care 3. Professor Saltman: "Accommodating Diversity in Swedish Health Provision The presentations will be followed by a general discussion on the topic of the session to bring out lessons for future research. Generally, we would be interested in understanding where the knowledge gaps are and how they can be addressed through research. 8

9 Temasession B: (Re)considering a coherent approach to operationalize an appropriate perspective in economic evaluation in Sweden Organisatör: Kasper Munk Johannesen, AstraZeneca och Linköpings universitet, Martin Henriksson, AstraZeneca Tid: Torsdag 20 mars 14:45-16:00 Moderator: Martin Henriksson, Astra Zeneca Medverkande: Kasper Munk Johannesen, AstraZeneca, Linköpings universitet, Martin Henriksson, AstraZeneca, Mark Sculpher, University of York, Joakim Ramsberg, Agency for Health and Care Analysis Introduction The underlying principles of economic evaluation in healthcare and the purpose of the healthcare sector, i.e. the objective function of the health care sector, provide the foundation for practical evaluations. However, whether we are to maximize health or utility, have a societal or budget optimization perspective, as well as other underlying principles are often taken for granted and are rarely scrutinized in great detail. This is somewhat problematic as the appropriate approach to performing economic evaluations only can be determined on the basis of these underlying principles. Recent debates about whether willingness-to-pay or the marginal productivity of the healthcare sector should form the basis for decision making, and whether future and non-healthcare costs should be included in cost-effectiveness analyses, clearly demonstrate a lack of consensus. The lack of consensus is likely a consequence of ambiguity around the underlying principles and objective function, different practical trade-offs and also incorporation of other implicit value judgments. Consensus may never be reached about the appropriate objective function of the health care sector. Revisiting the key value judgments and clearly outline how different objective functions leads to different conclusions on what should and should not be included in economic evaluations, seems to be key for having an informed debate around the use of economic evaluations is Sweden. Objective To have speakers outlining and discussing the key normative and practical considerations in economic evaluation. Speakers will discuss how different views on these issues leads to different conclusions regarding the appropriate approach to economic evaluation (e.g. what should be included in the evaluations and whether health or utility should be maximized). Presentations 1. Setting the scene Kasper Munk Johannesen / Martin Henriksson 2. Decision-maker perspective Mark Sculpher outlines the UK perspective and the considerations and debates that have taken place in the UK. 3. Welfarist (economics) perspective Joakim Ramsberg outlines one societal perspective operationalized through cost per QALY analysis. 4. General discussion moderated by the organizers Auditorium and speakers 9

10 Temasession C: Systematiska översikter av hälsoekonomiska utvärderingar Hur gör man och vad är det värt? Organisatör: SBU Tid: Fredag 21 mars 11:15-12:30 Moderator: Marianne Heibert Arnlind Medverkande: Emelie Heintz, Thomas Davidson, Harald Gyllensvärd, Marianne Heibert-Arnlind Systematiska översikter av hälsoekonomisk litteratur görs regelbundet i samband med utvärderingar av medicinska metoder. Syftet är då oftast att identifiera befintlig litteratur avseende olika metoders kostnadseffektivitet. För att göra en bra systematisk översikt behövs kunskap om vilka databaser som är relevanta att söka i, vilka sökstrategier och sökord som är lämpliga att använda, hur man granskar studiers relevans och kvalitet samt hur studier kan sammanställas på ett transparent sätt. SBU arbetar kontinuerligt med att utveckla och förbättra vårt arbetssätt inom dessa områden. Sammanställningar av kostnadseffektivitetsstudier kritiseras ibland för att de inte kan besvara frågan om en metod är kostnadseffektiv eller inte. Studierna som identifieras är ofta inte relevanta för den specifika kontexten, ibland baserade på gamla data eller med så stor variation i metoden att de inte blir jämförbara. Det finns emellertid andra syften med att identifiera vilka hälsoekonomiska studier som gjorts inom det utvärderade området. Några exempel är att motivera och hitta data till egna analyser, identifiera den mest relevanta studien för beslutsproblemet samt att hitta och förstå viktiga ekonomiska problem och samband kopplade till den specifika frågan. Syftet med denna temasession är att presentera hur översikter av hälsoekonomiska utvärderinger kan göras på ett systematiskt och transparent sätt samt att diskutera värdet av att göra denna typ av översikter. Vi kommer att presentera hur SBU arbetar med databaser, sökfilter, relevansgranskning, kvalitetsgranskning och presentation av resultat. Därefter diskuteras möjliga användningsområden och översikternas begränsningar. Presentation 1: Systematiska översikter av hälsoekonomiska utvärderingar Hur gör man? Presentation 2: När är det värt att söka systematiskt? Fördelar och nackdelar med systematiska översikter av hälsoekonomiska utvärderingar 10

11 Temasession D: Validering av hälsoekonomiska modeller varför, hur och för vem? Organisatör: Michael Willis, Institutet för hälso- och sjukvårdsekonomi, IHE, Lund Tid: Fredag 21 mars 11:15-12:30 Moderator: Sixten Borg, Lunds universitet, IHE och Adam Lundqvist, IHE. Medverkande: Michael Willis, IHE; Christian Asseburg, Esior OY; Martin Eriksson, TLV. Hälsoekonomiska simuleringsmodeller är viktiga verktyg för att beräkna kostnadseffektivitet inför införande av nya medicinska teknologier. Modeller passar särskilt väl för ex-ante beslutssituationer där simuleringar utifrån tillgängliga data kan ge vägledning om vilka faktorer som påverkar kostnader och hälsovinster på längre sikt. Exempelvis utgör modellanalyser en central del av beslutsunderlaget för myndigheter som Tandvårds- och läkemedelsförmånsverket, TLV, och dess brittiska motsvarighet NICE. Modelldesignen anpassas efter beslutsproblemet och konstruktören kan välja mellan enkla beslutsträd, markovmodeller och sekvensiella modeller (discrete event simulation). Även om en modell inte bör göras mer omfattande än nödvändigt, är sjukdomsförlopp många gånger komplexa och ett flertal samverkande riskfaktorer kan påverka utfall. Detta gäller inte minst förebyggande behandlingar som ska minska risken för framtida sjukdom och förtida död. Ett exempel är hjärtkärlsjukdom där det finns en mångfald förebyggande åtgärder liksom underliggande komplexa orsakssamband och där andra sjukdomar ingår bland riskfaktorerna. Inte minst när sjukdoms- och behandlingssamband är komplexa är validering ett viktigt led i utvecklingen av en modell. Det övergripande syftet med valideringen är att säkerställa att modellens kan generera trovärdiga resultat och beslutsunderlag. I takt med att modeller allt oftare byggs för en sjukdom eller ett sjukdomsområde i syfte att kunna analysera alla behandlingar för den aktuella patientgruppen, växer också behovet av att etablera principer för validering av modeller. 1. Hur kan hälsoekonomiska modeller valideras? Vilka aspekter är viktiga? Föredraget syftar till att ge en inledande bakgrund om valideringsprinciper och set the scene. PhD Michael Willis, Institutet för hälso- och sjukvårdsekonomi, IHE, Lund. Willis är ekonom och har mångårig erfarenhet av att utveckla hälsoekonomiska modeller, exempelvis inom diabetes. Han är medlem av det internationella MountHood samarbetet års erfarenhet av validering av diabetesmodeller det internationella Mount Hoodsamarbetet. Vad kan vi lära av det? Diabetes är en komplex sjukdom med ökad risk för att insjukna i bland annat hjärt-kärlsjukdom. Personer med diabetes behandlas förebyggande i syfte att förhindra diabeteskomplikationer många år framåt i tiden. Det är en av förklaringarna till att behovet av strategier för att validera diabetesmodeller tidigt varit tydligt. PhD Christian Asseburg, Esior OY. Asseburg är bayesiansk statistiker och har tillsammans med Willis utvecklat en mikrosimuleringsmodell för typ 1- och typ-2 diabetes. Han är medlem av det internationella MountHood samarbetet 3. Hur vill användare som TLV att modellen ska vara validerad? Vilka kriterier finns för att bedöma? Vilken typ av information förväntas modellkonstruktören redovisa från modellvalideringen. Martin Eriksson, hälsoekonom vid TLV. 4. Avslutande gemensam diskussion 11

12 MUNTLIGA FÖREDRAG Parallellsession A Tid: 20 mars 16:15-17:45 1. Forgetting to remember or remembering to forget -a study of the recall period length in health care survey questions Gustav Kjellsson a,b, Philip Clarke c, Ulf-G Gerdtham a,b,d adepartment of Economics, Lund University bhealth Economics & Management, Institute of Economic Research, Lund University c Centre for Health Policy, Programs and Economics, School of Population Health, The University of Melbourne dcenter for Primary Health Care Research, Lund University/Region Skåne Self-reported data on health care use is a key input in a range of studies. However, the length of recall period in self-reported health care questions varies between surveys, and this variation may affect the results of the studies. This article uses a large survey experiment to examine the role of the length of recall periods for the quality of self-reported hospitalization data by comparing registered with self-reported hospitalizations of respondents exposed to recall periods of one, three, six, or twelve months. Our findings have conflicting implications for survey design, as the preferred length of recall period depends on the objective of the analysis. For an aggregated measure of hospitalization, longer recall periods are preferred. For analysis oriented more to the micro-level, shorter recall periods may be considered since the association between individual characteristics (e.g., education) and recall error increases with the length of the recall period. 2. Short- and Long-Term Effects of Free Care Services for Mothers and Infants - Evidence from a Field Experiment Sonia Bhalotra a, Martin Karlsson b, Therese Nilsson c a University of Bristol b University of Duisburg-Essen c Lunds universitet Using information from a field trial in maternal and infant care running from 1931 to 1933 in seven Swedish medical districts, this paper tests the impact of ante- and neonatal care services on mothers and infants health. The arbitrary allocation of the programme to the districts and information to exactly identify eligible individuals (4,900 mothers and 5,000 children) gives a unique possibility to identify a causal relationship. We use two main data sources to perform two types of analyses. First, we collect aggregate information from annual reports from the 447 medical districts in Sweden, for the time period These reports include summary statistics on deliveries, the number of midwifes, child 12

13 mortality, complications, maternal mortality etc. Second, we use individual--- level data from church books, providing information on individuals being born, their sex, mothers marital status and subsequent mortality by death cause for mothers and infants, from the test districts for the years We compare outcomes in these districts to a set of control districts, which are chosen based on observable characteristics in the 1930 census and matched according to Mahalanobis distance. Our difference---in---differences estimates suggest both interventions had substantial effects, but on different outcomes. For example maternal care seems to have substantially reduced complications at delivery. Moreover, the effect on maternal mortality is considerable, albeit imprecisely estimated. As regards the effects on the mortality of the eligible children, we identify large and persistent effects of the infant care intervention, whereas we fail to find any impact of the maternal care intervention. We also find strong evidence of substantial treatment effect heterogeneity between different groups of the population: young mothers and single mothers appear to have been particularly responsive to the treatment. 3. Early-life exposure to adverse disease environment and labor force exit: Southern Sweden Tommy Bengtsson, Jonas Helgertz Centre for Economic Demography, Lund University A growing literature suggests that wellbeing and health in adult life is affected by exposure to malnutrition and disease during the fetal stage or in infancy. In this article we use longitudinal microlevel data for cohorts born between 1912 and 1924 in Southern Sweden to contrast the two hypotheses, the fetal origins hypothesis and the infancy inflammation hypothesis, aiming to estimate their respective relevance for the individual s decision to exit the labor force and into retirement. In testing the fetal origins hypothesis, we analyze, following the approach of Almond (2006), whether in-utero exposure to the 1918/1919 Spanish influenza pandemic is associated with a premature labor force exit, compared with surrounding non-exposed cohorts. In testing the infancy inflammation hypothesis, we focus on infant exposure to years characterized by substantially elevated rates of infant mortality, taking place in 1913 and 1916 due to pneumonia and whooping cough, respectively. In a previous study (Bengtsson and Helgertz, forthcoming), we have examined the labor force entry and early career of individuals exposed to the Spanish flu while in-utero or to the adverse disease environment of 1913/1916 during infancy. In less than fully adjusted models, we find point estimates from in-utero exposure to the Spanish flu that are highly similar to the results of Almond (2006). Albeit statistically insignificant, this we to a considerable extent attribute to the substantially smaller sample size of our study. In fully extended models, also controlling for the changing macroeconomic environment over time, the effect of in-utero exposure to the Spanish flu becomes modest in size, remaining not significant. Instead, we find negative income effects of exposure to an adverse disease environment, in particular to pneumonia and whooping cough, during the first years of life. In terms of size, the effect is statistically significant as well as more than half that of the difference between high and low socioeconomic origins. 13

14 We have found few indications that would suggest that either exposure results in an inability to enter the labor market. Thus, the mechanisms are tentatively more likely to be represented by a suboptimal performance in the labor market, potentially through a higher incidence of absence due to disability or sickness. The proposed study extends on our previous study, examining whether the suggested mechanisms accumulate over the life course, resulting in an earlier withdrawal from the labor force, or whether they are direct. References Almond, D Is the 1918 Influenza Pandemic Over? Long-term effects of In Utero Influenza Exposure in the Post-1940 U.S. Population. Journal of Political Economy, 114: Bengtsson, T., Helgertz, J. Effects of early-life exposure to the Spanish flu and adverse disease environment on adulthood incomes: Southern Sweden The impact of a childhood onset of Type 1 diabetes on higher education and labour market outcomes Sofie Persson 1), Gisela Dahlquist 2), Ulf. G. Gerdtham 1,3), Katarina Steen Carlsson 1) 1) Health Economics Unit, Department of Clinical Sciences, Lund University, Malmö 2) Pediatrics Unit, Department of Clinical Sciences, Umeå University 3) Economics Department, Lund University In this study we investigate the impact of a sudden health shock in childhood on labour market outcomes in early adulthood. We use the childhood onset of type 1 diabetes as an example of an exogenous health shock with direct negative health effects for the child. Individuals, diagnosed <15 years and born in , were selected from the Swedish Childhood Diabetes Register, which was linked to national population registers including the Longitudinal integration database for health insurance and labour market studies (LISA). For each individual (n=2,485), four controls from the general population, matched for year of birth and residency at the time of the diagnosis, were selected by Statistics Sweden (n=9,940). The impact of diabetes on educational level, employment and earnings was analysed both at the age of 32 and throughout ages The analyses show a negative impact of diabetes on the level of education and labour market outcomes among young adults. Women with the disease were generally worse off and the impact appeared to increase with the duration of diabetes. In conclusion, the results from this study suggest that an early onset of a chronic disease, such as type 1 diabetes, negatively impacts the future labour market status of the affected individuals. 14

15 Parallellsession B Tid: 20 mars 16:15-17:45 1. En utvärdering av det statliga tandvårdsstödet rapport från TLV:s regeringsuppdrag Douglas Lundin TLV När det statliga tandvårdsstödet infördes 2008 bestod det av två delar: ett allmänt tandvårdsbidrag (ATB) och ett högkostnadsskydd. Förra året tillkom ett särskilt tandvårdsbidrag. Vi diskuterar framförallt tre frågor: Finns det stora skillnader mellan olika inkomstgrupper vad gäller hur mycket tandvårdsstöd de tar emot? Har priskonkurrensen på tandvårdsmarknaden förändrats sedan stödet infördes? Hur påverkar referenspriserna de priser som tandvårdsmottagningarna sätter? Den första frågan motiveras av att syftet med högkostnadsskyddet är att möjliggöra för individer med stora tandvårdsbehov att få tandvård till en rimlig kostnad. Det kan finnas olika anledningar till att individer med behov av tandvård väljer att inte åtgärda sina problem: rädsla, att man prioriterar annat, dålig ekonomi, etc. Men det är bara en av dessa faktorer som högkostnadsskyddet kan påverka: dålig ekonomi. Om detta är en vanlig orsak till att individer inte åtgärdar sina tandproblem, då borde det återspeglas i att låginkomsttagare i mindre utsträckning åtgärdar sina problem än medel- och höginkomsttagare. Den generella slutsatsen är att det inte finns några stora skillnader mellan olika inkomstgrupper i mottaget stöd. Den andra frågan, om priskonkurrensen har förändrats, motiveras av att en målsättning med stödet är att det ska utformas så att det främjar en positiv utveckling av tandvårdsmarknadens funktionssätt. Stödet ska vara en subvention av patienternas pris och inte en subvention till vårdgivarna. Vår slutsats är att priskonkurrensen inte har förbättrats. Vårdgivarpriserna har ökat snabbare än den allmänna prisnivån i svensk tjänsteproduktion och även snabbare än tandvårdens referenspriser (som ska återspegla kostnaderna). Den tredje frågan, hur referenspriserna styr vårdgivarnas priser, motiveras av att ett av syftena med att införa referenspriser var att de skulle underlätta tandvårdskonsumentens jämförelser av priser. Vi konstaterar först att referenspriserna styr vårdgivarnas priser mycket tydligt: två av tre patienter möter ett vårdgivarpris som maximalt ligger 10 procent över referenspriset. Orsaken till sambandet är att referenspriserna styr vårdgivarnas priser på ett direkt sätt, snarare än att det är en tredje faktor produktionskostnaderna som påverkar båda. En av indikationen på detta är att vårdgivarna ändrar sina priser i samband med referensprisändringar även i de fall där ändringarna inte berott på ändringar i produktionskostnaden. Att referenspriserna styr vårdgivarpriserna direkt är inte förvånande och behöver inte heller vara ett problem. Men det ställer krav på TLV att göra bra uppskattningar av produktionskostnaderna så att referenspriserna hamnar på en rimlig nivå. 15

16 2. Towards a consistent use of Cost of Added Life-Years by TLV Audun Ohna GSK Sweden Background: The general guidelines (LFNA 2003:2) on economic assessments recommend that the health economic evaluation of pharmaceuticals should have a societal perspective, both including all relevant healthcare and non-healthcare related consequences. Societal perspective in economic evaluation of pharmaceuticals is not uncommon, however, including non-healthcare related consumption due to survival improvements is controversial. TLV is one of the few reimbursement authorities, if not the only, to enforce the inclusion of cost of added life-years in the base case assessment of pharmaceuticals with expected survival benefit. Objective: The objective of this work was to evaluate the consistency of the consideration of cost of added life-years of pharmaceuticals with expected survival benefit, in the reported TLV assessments from Klinikläkemedelsprojektet Method: All TLV assessment reports from the Klinikläkemedelsprojektet were reviewed. Reports assessing pharmaceuticals with only quality-of-life benefits were excluded. Characteristics of the economic evaluation assessed were recorded, in addition to cost of added life-years. Results: By , thirteen assessments included pharmaceuticals with expected survival benefit. Approximately half of the reports were related to cancer; about 20% related to hematology; and 15% related to autoimmune diseases. The mean or median age of the patient population in the assessments ranged from 20 to 69 years, however, for cancer the age of the patients ranged from 54 to 69 years. Eleven of these assessments included cost-effectiveness analysis, of which ten included calculations of indirect costs. TLV considered cost of added life-years in seven of these assessments. Discussion and conclusion: In almost all of the reviewed assessments, TLV has considered indirect costs. From 2012 to 2013, there has also been a clear trend, where TLV has become more transparent and more explicit in reporting indirect costs in the reports from Klinikläkemedelsprojektet. Four of the assessments relates to populations above an age of 65, which implies that these patients are net consumers. In these cases, societal perspective results in a higher cost per QALY gained. One implication of this praxis, although it is not intentionally its purpose, could be less prioritization to patients outside the working age population. This work demonstrates that TLV is approaching a more transparent, explicit and consistent enforcement of cost of added life-years, where every assessment is considered by the same conditions. This risk is, nevertheless, that we end up with consistent decisions based on wrong conditions. 16

17 3. Betydelsen av kostnadseffektivitet och sjukdomssvårighet för subvention av läkemedel i Sverige: en analys av Tandvårds- och läkemedelsförmånsverkets beslutsfattande Mikael Svensson a, Fredrik Nilsson b, Karl Arnberg b a) Karlstads universitet b) Tandvårds- och läkemedelsförmånsverket Abstrakt I den här studien analyserar vi hur kostnadseffektivitet och sjukdomssvårighet påverkar subventionsbeslut av läkemedel för Tandvårds- och Läkemedelsförmånsverket (TLV). Kostnadseffektivitet mäts genom kostnaden per QALY medan sjukdomens svårighetsgrad baseras på en bedömning av TLV:s sakkunniga råd huruvida läkemedlet rör en sjukdom med hög svårighet eller ej. Vi analyserar alla subventionsbeslut från år 2005 till 2011 där det finns data om kostnaden pe QALY och sjukdomens svårighetsgrad. Totalt ger detta 102 olika beslut där 86 beviljade subvention och 16 nekade subvention. Mediankostnaden per QALY för läkemedel som beviljades subvention var kronor medan mediankostnaden per QALY för läkemedel som nekades subvention var kronor. Av de beviljade subventionerna varierar kostnaden per QALY från <0 upp till kronor, medan nekade subventioner har en kostnad per QALY från kronor upp till kronor. Resultat från en logistisk regressionsanalys visar att kostnaden per QALY och sjukdomens svårighetsgrad både är statistiskt signifikant relaterade till sannolikheten för beviljad subvention. När kostnaden per QALY överstiger kronor (ej svåra sjukdomar) och kronor (svåra sjukdomar) är det mer sannolikt att subvention nekas än att subvention beviljas. I Sverige används ibland en tumregel om kronor per QALY som en gräns för en kostnadseffektiv åtgärd. Vår modell visar dock att vid denna kostnadseffektivitet är sannolikheten för beviljad subvention av ett läkemedel procent. 4. Hälsoekonomiska analyser för insatser till sköra äldre SBU:s projekt inom äldreområdet Thomas Davidson, Harald Gyllensvärd SBU Nuvarande demografiska utveckling innebär att antalet och andelen äldre kommer att öka i framtiden. Detta ställer stora krav på vården och omsorgen att, i de fall det är möjligt, förebygga och behandla olika sjukdomar hos äldre och sköra äldre. Av den anledningen har Statens beredning för medicinsk utvärdering (SBU) tillsammans med Socialstyrelsen fått ett regeringsuppdrag att utvärdera flera metoder för vård av sköra äldre. En del i dessa utvärderingar är att studera de hälsoekonomiska aspekterna. Behandling av urininkontinens samt hur akutvården ska organiseras är två utvärderingar som SBU nyligen publicerat. Därutöver pågår projekt om nytta-/riskindikatorer för vanliga hjälpmedel, kosttillägg vid undernäring, prevention och behandling av svårläkta sår samt behandling av depression. 17

18 Det finns särskilda hälsoekonomiska utmaningar när insatser till gruppen sköra äldre ska utvärderas. Bland annat finns det generellt få studier på äldre individer ( 65 år) och än färre på sköra äldre. I brist på studier är det ofta svårt att påvisa att en behandling är effektiv. Likaså finns svårigheter med att det ofta är en heterogen patientgrupp med olika komorbiditeter. För patienter med ett flertal sjukdomar kvarstår ofta vårdbehov även om den primära diagnosen behandlas. Inom hälsoekonomin finns metoder för att analysera förväntade konsekvenser vid omfattande osäkerhet, men om det helt saknas kliniska studier, eller om dessa inte är relevanta för en svensk kontext, är det svårt att göra relevanta hälsoekonomiska beräkningar. I rapporten Behandling av urininkontinens hos äldre och sköra äldre framkommer det att en halv miljon äldre i Sverige lider av urinläckage. Urininkontinens påverkar livskvaliteten negativt och medför stora totala kostnader som till stor del kan hänföras till hjälpmedel och merkostnader inom äldrevårdens olika boendeformer. Det saknas emellertid studier av tillräckligt god kvalitet inom området för att fastställa olika insatsers kostnadseffektivitet, men två insatser som i rapportens diskussion förefaller ha en rimlig kostnadseffektivitet för äldre är bäckenbottenträning samt operation med intravaginal slyngplastik. I rapporten Omhändertagande av äldre som inkommer akut till sjukhus med fokus på sköra äldre visas bristfällig relevans hos det vetenskapliga underlaget. Därför gjordes egna beräkningar avseende resursförbrukning. Beräkningarna visade att integrerat strukturerat omhändertagande av äldre leder till en direkt ökad kostnad på cirka kronor per patient. Metoden kan dock leda till samhälleliga kostnadsbesparingar eftersom fler patienter kan bo kvar hemma samt att färre patienter behöver återinläggas på sjukhus. Sammantaget är det rimligt att anta att integrerat strukturerat omhändertagande av äldre är kostnadseffektivt i jämförelse med vanlig vård. 18

19 Parallellsession C Tid: 20 mars 16:15-17:45 1. Is the level of patient co-payment for medicines associated with refill adherence in Sweden? Eva Lesén1,2, Karolina Andersson Sundell1,3, Anders Carlsten4, Ann-Charlotte Mårdby3,5, Anna K Jönsson6 1. Nordic School of Public Health, Gothenburg 2. Nordic Health Economics AB, Gothenburg 3. Social Medicine, Institute of Medicine, University of Gothenburg 4. Medical Products Agency, Uppsala 5. Analysis Unit, Sahlgrenska University Hospital, Gothenburg 6. Department of Drug Research/Clinical Pharmacology, Faculty of Health Sciences, Linköping University / Department of Clinical Pharmacology, County Council of Östergötland, Linköping ABSTRACT BACKGROUND: In the Swedish reimbursement scheme, the co-payment is based on the price of the product and decreases in a stepwise manner as the total accumulated co-payment increases. The aim of this study was to analyse how refill adherence in Sweden varies according to the patient's copayment level for medicines, with antiepileptic drug (AED) use as an example. METHODS: Prevalent AED users aged years who purchased an AED between 1 January and 30 June 2007 were identified in the Swedish Prescribed Drug Register and followed for a maximum of 2 years. Patient time was categorized based on the patient's accumulated co-payment for all drugs per reimbursement period. The continuous measure of medication acquisition (CMA) was used to estimate refill adherence in relation to the patients' co-payment level. Associations between patients' co-payment for all medicines and refill adherence were assessed with multilevel mixedeffects linear regression, accounting for clustering within patients. RESULTS: The study population included 2210 patients (mean age: 56 years; 54% men). CMA for AED was 91% for patients where the co-payment corresponded to 100% of the price. Compared with these patients, refill adherence for AED was 2-4% higher (P < 0.001) for patients with reduced copayment (co-payment of 50% of the price). Higher age, higher income and fenytoin use were also associated with a higher refill adherence for AED. CONCLUSIONS: Using AED as an example, a higher level of reimbursement was associated with a higher refill adherence compared with full co-payment in Sweden. 2. Pharmaceuticals An analysis based on Swedish real-life data Sofie Gustafsson Lunds universitet Low patient adherence to prescribed pharmaceutical therapies attenuates treatment benefitsand impedes health care effectiveness. This widespread and persisting problem is a public concern attracting global attention from policy makers and health managers. Despite a large literature, the 19

20 mechanisms determining the individual s decision process regarding adherence behavior is not yet well understood. This study analyses patient adherence to prescribed pharmaceuticals within the demand-forhealth model. Despite being fundamental for economic analyzes for individual health behavior, the model has not earlier been used to explain patient adherence to pharmaceuticals. This paper adds two features to the original model. First, it assumes that the prescriptions are correctly prescribed so that adherence follows the physiological optimal quantity constraint. This implies that any deviation from the prescribed treatment plan is penalized by generating less health investments than feasible. Second, it assumes that the rate at which health investments transform to actual health depends on the characteristics of the specific pharmaceutical (i.e. its ability to interact and modify human biological systems) and patient characteristics, which determine the specific patient s biological response to treatment. Besides characteristics such as gender and general health, the manifestation of specific health conditions may affect the patient s biological response to pharmaceutical treatment making adherence more or less important. Thus, by modifying the health penalty size for deviating, the specific health conditions change the health benefits of adhering. For cardiovascular disease management, a previous manifested heart disease or a prolonged hypertension spell are specific health conditions augmenting the health penalty size for deviating from prescribed pharmaceuticals with selective beta receptor blocker, ace inhibitors, angiotensin receptor blockers and diuretics. Correspondingly for the treatment of mental illnesses, previous manifestation of anxiety or depression augments the health penalty size for deviation from antidepressants. As a higher a health penalty implies a higher marginal health benefit when adhering, theory predicts that these specific health conditions increase adherence to the corresponding regimens. For the empirical analysis, the paper uses a dataset containing the year s wave of Swedish Survey of Living Conditions (ULF) merged with the Swedish Prescribed Pharmacy Register (SPPR) from July 2005 through November Consisting of a national representative sample of approximately 10,000 adults, respondents in ULF were interviewed about living conditions, health and socioeconomic circumstance and answers were complemented with register data on, for instance, income, tax transfers and in-patient hospitalization. The nation-wide SPPR register contains information on all pharmacy dispensed pharmaceuticals with prescription and includes the patient s unique prescribed daily dosage regimen. This data enabled us to assess 18 month adherence to a set of pharmaceuticals commonly used in the long-term treatment of mental illness and cardiovascular diseases. Controlling for general health, socioeconomic and demographic factors, the empirical analyses show that specific health conditions associate positively with adherence. The positive association implies that individuals deviate less to prescribed pharmaceutical treatments when the marginal health benefit of adherence is higher. This finding is important when designing interventions aiming at improving long-term adherence to pharmaceutical regimens. 20

21 3. A new approach to decomposing the concentration index of health using the recentered influence function Gawain A Heckley1,2, Ulf-G Gerdtham1,2,3 1Health Economics & Management, Institute of Economic Research, Lund University 2Division of Health Economics, Department of Clinical Sciences, Malmö, Lund University 3 Department of Economics, Lund University This paper introduces a linear approximation decomposition technique applicable to various forms of the Concentration index of health. The technique consists of two stages: the first is to estimate the first order linear approximation of the Concentration index of health a Recentered Influence Function (RIF) which is a transformation of the y variable; the second stage is to integrate up the conditional RIF yielding the Concentration index by performing a regression using the RIF as the dependent variable a RIF regression. Assuming linearity in parameters allows OLS regression of the RIF, yielding the decomposition estimates. This approach has several advantages over current practice based upon a linear regression of health: It captures the effect of the covariates on both health and rank, the same technique can be performed on all types of health variable allowing consistency of methodology and statistical inference is much simpler. However, it is a linear approximation and is therefore only appropriate for relatively small changes. This limitation depends on the empirical context. The method is applied to income related self-reported health inequality in Sweden and compared to current practice as an illustration. 4. Income receipt and mortality - evidence from Swedish public sector employees Elvira Andersson Lunds universitet A large literature has established a positive long-run relationship between health and income, with lower mortality and morbidity rates for higher-income individuals. However, in developed countries mortality rates follow a pro-cyclical pattern, suggesting that this positive association does not apply to temporary income changes on an aggregate level (Ruhm 2000, Neumayer 2004, Tapia Granados 2005). A possible explanation is that income receipt has adverse short-run health effects that partly offset the positive long-run income/health association due to an increase in activity upon income receipt. In this paper, we address the relationship between income and mortality by studying the short-run association between periodic and expected income receipt and death rates. Using a panel combining register data and survey information on exact pay-days for each individual employed by the Swedish public sector during , we estimate the effect of pay-check receipt on mortality. Variation in pay-days across work-places allows us to completely control for general within-month and withinweek mortality patterns and mortality patterns related to e.g. holidays and special events coinciding with pay-days. The paper addresses the mortality channels by linking information on causes of death to each deceased individual. 21

22 Our sample corresponds to approximately 30% of the Swedish work force and is heterogeneous in age, income and education. This allows us to both study a representative population and investigate differences between sub-groups. By comparing mortality responses between income quartiles, we study the role of liquidity constraints. We also compare the effects between men and women and different-age individuals to investigate differences in behavior and health. The relationship between income receipt and short-term mortality has previously been studied by Evans and Moore (2011), who find a substantial increase in mortality following income receipt (i.e. pay-checks and social security checks) amongst military personnel and the elderly. We go beyond this study by using an identification strategy which credibly controls for both cyclical and non-cyclical variations in mortality and by studying and quantifying mortality effects for both a representative population and several subgroups. We find a dramatic increase in total mortality on the day pay-checks arrive. The increase is especially pronounced for young individuals and for deaths due to circulatory conditions, e.g. heart conditions and strokes. This is consistent with an increase in general activity being an important cause of the excess mortality. The effect is entirely driven by lower income individuals, suggesting that liquidity constraints are an important factor behind the mortality response. 22

23 Parallellsession D Tid: 21 mars 08:15-09:45 1. SWEDISH EXPERIENCE-BASED VALUE SETS FOR EQ-5D HEALTH STATES Kristina Burström1, 2, 3, Sun Sun 1, 3, Ulf-G Gerdtham 4, 5, 6 Martin Henriksson, 7 Magnus Johannesson,8 Lars-Åke Levin 9, Niklas Zethraeus 1, 10 1 Medical Management Centre, Department of Learning, Informatics, Management and Ethics, Karolinska Institutet 2 Division of Social Medicine, Department of Public Health Sciences, Karolinska Institutet, 3 Stockholm County Council, Health Care Services 4 Department of Economics, Lund University 5 Health Economics & Management, Institute of Economic Research, Lund University 6 Centre for Primary Health Care Research, Lund University 7 Department of Health Economics, AstraZeneca Nordic 8 Department of Economics, Stockholm School of Economics 9 Center for Medical Technology Assessment, Department of Medical and Health Sciences, Linköping University 10 The Dental and Pharmaceutical Benefits Agency ABSTRACT Purpose: To estimate Swedish experience-based value sets for EQ-5D health states using general population health survey data. Methods: Approximately 45,000 individuals valued their current health status by means of timetrade off (TTO) and visual analogue scale (VAS) methods and answered the EQ-5D questionnaire, making it possible to model the association between the experience-based TTO and VAS values and the EQ-5D dimensions and severity levels. The association between TTO and VAS values and the different severity levels of respondents answers on a self-rated health (SRH) question was assessed. Results: Almost all dimensions (except usual activity) and severity levels had less impact on TTO valuations compared with the UK study based on hypothetical values. Anxiety/depression had the greatest impact on both TTO and VAS values. TTO and VAS values were consistently related to SRH. The inclusion of age, sex, education and socioeconomic group affected the main effect coefficients and the explanatory power modestly. Conclusions: A value set for EQ-5D health states based on Swedish valuations has been lacking. Several authors have recently advocated the normative standpoint of using experience-based values. Guidelines of economic evaluation for reimbursement decisions in Sweden recommend the use of experience-based values for QALY calculations. Our results that anxiety/depression had the greatest impact on both TTO and VAS values underlines the importance of mental health for individuals overall HRQoL. Using population surveys is in line with recent thinking on valuing health states and could reduce some of the focusing effects potentially appearing in hypothetical valuation studies 23