Many countries are experimenting with new models to better integrate care; yet, innovative care models are often implemented as time-limited, localised projects with limited impact on service delivery more broadly. This paper seeks to understand the processes behind successful projects that achieved some form of 'routinisation' and informed systemwide integrated care strategies. It draws on detailed case studies of three integrated care experiments: the 'Integrated effort for people living with chronic diseases' project in Denmark; the Gesundes Kinzigtal network in Germany; and Zio, a care group in the Maastricht region in the Netherlands. It explores how they were developed, implemented and sustained, and how they impacted the wider system context. All three models implicitly or explicitly adopted processes shown to be conducive to the dissemination of innovations, including dedicated time and resources, support and advocacy, leadership and management, stakeholder involvement, communication and networks, adaptation to local context and feedback. Each showed robust evidence of improvements on a number of service and patient outcomes and these findings were central to their wider impacts, shaping country-wide integrated care polices. However, the wider dissemination of projects occurred in an incremental and somewhat haphazard way. To further redesign health and social care a more formal strategy, alongside resources, may thus be needed to provide funders and providers with genuine incentives to invest in new business models of care. There remains a crucial need for better understanding of specific local conditions that influence implementation and sustainability to enable translation to other contexts and settings.

Phone: +49 40 22621149-0 page 4 Conflict of interestMr. Schulte, Dr. Groene and Mr. Hildebrandt are employees of the OptiMedis AG. Dr. Pimperl was also employed by the OptiMedis AG, but is currently a Harkness Fellow in Health Care Policy and Practice and does not receive any funding from the OptiMedis AG. The OptiMedis AG is the management partner and shareholder of the Gesundes Kinzigtal GmbH, which is used as study setting for the empirical application of the evaluation design described in this paper. Mr. Hildebrandt is also the CEO of the Gesundes Kinzigtal GmbH. There are no conflicts of interest for the other co-authors. FundingAt the time this manuscript was developed, Dr. Pimperl was engaged in the Harkness Fellowship in Health Care Policy and Practice, supported by The Commonwealth Fund, a private independent foundation based in New York City. The views presented here are those of the authors and not necessarily those of The Commonwealth Fund, its directors, officers, or staff.page 5 Abstract:A central goal of accountable care organizations (ACOs) is to improve the health of their accountable population. No evidence currently links ACO development to improved population health. A major challenge to establishing the evidence-base for the impact of ACOs on population health is the absence of a theoretically-grounded, robust, operationally feasible, and meaningful research design. We present an evaluation study design, provide an empirical example, and discuss considerations for generating the evidence-base for ACO implementation.A quasi-experimental study design using propensity score matching in combination with smallscale exact-matching is implemented. Outcome indicators based on claims data were constructed and analyzed. Population health is measured by using a range of mortality indicators: mortality ratio, age at the time of death, years of potential life lost / gained and survival time. The application is assessed using longitudinal data from Gesundes Kinzigtal, one of the leading population-based ACOs in Germany.The proposed matching approach resulted in a balanced control of observable differences between the intervention (ACO) and control group. The mortality indicators used, indicate positive results. For example, 635.6 fewer years of potential life lost (2,005.8 vs. 2,641.4; T-Test: sig. p<0.05*) in the ACO-intervention group (n=5,411), attributable to the ACO, also after controlling for a potential (indirect) immortal time bias by excluding the first half year after enrollment from the outcome measurement.Our empirical example of the impact of a German ACO on population health can be extended to the evaluation of ACOs and other integrated delivery models of care.

Regarding population health, most of the quality indicators examined by the external scientific evaluation show positive development. For example, the prevalence of patients with fractures among all insurants with osteoporosis is presented. In 2011, this prevalence was approximately 26 % in the "Kinzigtal" population (aged ≥ 20 years old) in comparison to 33 % in the control group. As far as patient experience is concerned, to the question "Would you recommend becoming a member of Gesundes Kinzigtal to your friends or relatives?" 92.1 % of those questioned answered "Yes, for sure" or "Yes, probably." Twenty-four percent of those questioned further stated that they would now live "more healthy" than before enrolment in the ICSGK. In the subgroup of questioned insurants who had objective agreements with their doctors 45.4 % answered in this way. On the subject of cost-effectiveness, for both participating socil health insurance schemes, cost savings relative to the costs normally expected for the ICSGK population concerned are observed every year. In the seventh intervention year (2012) the total is 4.56 million Euros for the AOK Baden-Württemberg (BW), which is a contribution margin of 146 Euros per insurant for the 31.156 insurants concerned (LKK BW = 322 Euros per insurant relative to cost savings). The results presented in this paper indicate positive effects in all three Triple Aim dimensions. Further longitudinal studies are recommended to validate those first results together with a detailed analysis to obtain in-depth insights into the specific influence of subcomponents of the total intervention.

Background Across countries, a small group of patients accounts for the majority of health care spending. These patients are more likely than other patients to experience problems with quality and safety in their care, suggesting that efforts targeting efficiency and quality among this population might have significant payoffs for health systems. Better understanding of similarities and differences in patient characteristics and health care use in different countries may ultimately inform further efforts to improve care for HNHC patients in these health systems. Methods We conducted a cross-sectional descriptive study using one year of patient-level data on high-cost patients in seven high-income OECD member countries. Countries were selected based on availability of detailed information (large enough samples of claims, administrative, and survey data of high-cost patients). We studied concentration of spending among high-cost patients, characteristics of high-cost patients, and per capita spending on high-cost patients. Findings Cost-concentration of the top 5% of patients varied across countries, from 41% in Japan to 60% in Canada, driven primarily by variation in the top 1% of spenders. In general, high-cost patients were more likely to be female (57.7% on average), had a significant number of multi-morbidities (up to on average 10 major diagnostic categories (ICD chapters), and had a lower socioeconomic status. Characteristics of high-cost patients varied as well: median age ranged from 62 in the Netherlands to 75 in Germany and the difference in socioeconomic status is particularly stark in the US. Lastly, utilization, particularly for inpatient care, varied with an average number of inpatient days ranging from 6.6 nights (US) to 97.7 nights in Japan. Interpretation In this descriptive study, there is substantial variation in the cost concentration, characteristics, and per capita spending on high-cost patient populations across high-income countries. Differences in the way that health systems are structured likely explains some of this variation, which suggests the potential of cross-system learning opportunities. Our findings highlight the need for further studies including comparable performance metrics and institutional analysis.