New Frontiers In Health Policy Evaluation European Health Policy Group (EHPG)
April 29-30th 2010

Zeynep Or







Assessing new models of hospital governance in Europe: how much real change?

Richard Saltman, Antonio Duran and Hans Dubois

Over the past 15 years, many tax-funded health systems across Europe have introduced some form of hospital self-management for publicly owned hospitals. All hospitals in the UK, Norway, and Estonia, as well as some hospitals in Spain, Portugal, Italy, and Czech Republic have been re-structured into quasi-independent self-managed institutions. Among other changes, hospitals have been weaned from fixed public budgets and required to obtain some or all of their funding through contract-based arrangements. Many have also been given new, publicly appointed but independently established Boards of Trustees. In many cases, hospital management also has been given increased autonomy to hire and fire staff, to determine the mix of services they provide, and, in certain instances (for example Foundation hospitals in the UK), to make investment decisions, borrow funds, and to decide about the disposition of buildings and real estate. In Israel, Ministry-owned hospitals have been allowed to compete in the private medical market for 30% of their total revenues, and to keep internally any earned surplus.

Despite all this formal change, however, many questions remain about how these new hospital governance models work in practice, and the degree of real change. This paper presents initial results from a European Observatory study of eight country cases, seeking to determine how much change has actually occurred. The first part develops a framework for comparing the degree of similarity among the eight countries' health governance systems, and the nature of the commonalities and differences in these new arrangements. The second part assesses the extent to which the day-to-day behavior of these reformed hospital governance systems actually reflects these new models – in other words, the degree to which new hospital governance decisions differ in practice from the prior, directly managed arrangements. It also, to the extent possible, examines the degree to which these more independently managed institutions have better levels of clinical, financial, and patient satisfaction performance.

Improving cancer detection, treatment and survival through new governance

Thomas Oliver, Louise Trubek, Toby Campbell, Matt Mokrohisky and Chih-Ming Liang

The Obama administration's commitment to a renewed campaign against cancer provides a timely platform to examine how new forms of regulation and governance can be utilized to improve cancer detection, treatment and survivorship. “New governance” is a regulatory model that differs from top-down, command and control regulation and from professional self-regulation. Instead, it relies on effective collaboration from multiple stakeholders, and emphasizes improved efficiency and innovation through data collection, monitoring and benchmarking. This paper discusses the emergence of new resources and strategies in the U.S. campaign against cancer as well as recent cancer initiatives in the European Union, and investigates the potential benefits and challenges associated with their implementation. In particular, we explore the role of new evidence, new technology, new participation and new institutions. For example, we examine the conditions under which new guidelines for cancer detection and treatment are adopted and implemented into standard bedside practice, and we investigate the capacity of voluntary networks such as the Wisconsin Collaborative for Healthcare Quality and the National Quality Forum in the U.S., as well as participants in the Europe Against Cancer program, to facilitate this process. In addition, we consider how the broader application of electronic medical records can strengthen analysis and communication among patients, clinicians, researchers and policy makers. We also investigate the comparative incentives and results of using public reporting of cancer outcomes versus direct financial incentives for improved performance. Finally, we examine how increased participation by patients, their families and advocacy groups can be translated into a new culture of clinical research and practice, resulting in expanded cancer registries, greater participation and representation in clinical trials, and improved focus on follow-up care and quality of life for cancer survivors.

Concentration of services for thoracic surgery: a register-based study of changes in quality and efficiency

Emma Medin, Pia Maria Jonsson, Torbjorn Ivert, Clas Rehnberg

Background Over the last decades health service researchers have found considerable evidence that the volume of hospital services is positively related to patient outcomes for certain diagnoses and surgical procedures. While this relationship is by no means universal it has been observed for a variety of cases ranging basic procedures as appendectomies to highly specialized care and quaternary care such as cardiac surgery and transplantation surgery. There are three major factors responsible for this volume–outcome relationship: skill of the surgeon, experience of the supporting team (anesthesiologists, intensivists, and nurses), and general resources of the hospital. Some studies have stressed the importance of the surgeon, whereas others have shown that hospital volume is the most important factor. Skilled surgeons tend to develop busy practices at high-volume tertiary care hospitals, so it is difficult to study the various factors in isolation. The relationship between the volume of hospital services and reduction of unit costs (economies of scale) has not been subject to as many studies as the volume-outcome relationship. The neo-classical theory assumes a U- shaped relationship between average unit cost and quantity and quality of output but this relation has not been easy to prove in empirical studies.

Objective Based on expectations of economies of scale and higher quality outcome two thoracic surgery departments was merged in 2004 as part of a large university hospital merger in Stockholm. The aim of this study is to examine whether this merger lead to reductions in unit costs and/or higher patient quality outcomes due to a hospital-volume outcome relationship, a surgeon-volume outcome relationship or due to specific patient related factors.

Data and methods Data from all thoracic surgery departments in Sweden (n=8) covering two years pre-merger and two years post-merger (2002 to 2006) is collected from different sources. Patient outcome data is gathered from the Patient Register, the National Quality Register for Cardiac Surgery and from the register of Swedish Patient Insurance. Cost data is collected from the cost per patient database at the Swedish Association of Local Authorities and Regions. The study design is a retrospective longitudinal top- down analysis based on register data. A multilevel regression analysis with three levels (hospital, surgeon and patient) is performed and the non-merged hospitals act as a control group.

Results, conclusion and policy implications TBC

Policies for health care cost control

Joe White

This paper will assess a wide range of policies in three ways. First, it creates a path analysis of health care costs and so of opportunities for policy interventions. From a policy perspective, downstream interventions - interventions closest to the final point of payment - are more effective, because more certain, than attempts to shape "fundamental" causes. For instance, measures to improve population health will still only reduce spending if factors further down the path - such as other factors which shape utilization, prices, and overhead - are managed appropriately. Capacity restrictions save money if prices (further downstream) are set in a way that doesn't translate restricted capacity into higher prices (i.e., sellers do not have market power because prices are regulated).

Second, the paper argues that there are three factors that shape policy success: knowledge (or technical capacity), power (or institutional capacity) and will (or political capacity). In order to understand the prospects for any given method, one has to assess it and those who will have to create and implement that method on all three dimensions. For example, it is relatively easy, technically and institutionally, to regulate prices, as opposed to trying to make individual episodes of care more appropriate. Both are politically difficult, but the former is probably easier than the latter, even in the United States. Third, the paper assesses the evidence about various approaches, using available studies from the U.S. and other advanced industrial nations. The paper therefore hope to provide an overall framework for thinking about a core health policy issue; the framework is original and based on observations about policy success and failure that apply well beyond health care policy.

Understanding complex system reform in the English national health service: a realist evaluation

Ross Millar, Martin Powell, Abeda Mulla and Anna Dixon

Background Recent health reform in England has sought to drive improvements across healthcare through supply, demand, transactional and system management levers and incentives. Such complex policy reform processes present a challenge to evaluation since it is essential to understand both the impact in terms of goals achieved and the process by which any observed benefits were accomplished.

Aims The paper presents a research framework based on Pawson and Tilley's realistic evaluation model (1997) to investigate the ‘causal factors' that influence how organisations and individuals respond to interventions that seek to change their behavior and impact on health outcomes. The paper considers complex policy reform by examining how they are appropriate in three tracer conditions - diabetes, orthopaedics and early intervention mental health.

Methodology The paper presents findings from a comparative case study design of six case study sites in England. It draws on semi-structured interviews (n=202) with a range of actors to assess the impact of reforms and how key mechanisms (and the inter-play between) act to enable or prevent the achievement of local objectives in tracer conditions.

Findings The research illustrates the variable impact of reform across tracer conditions. The reforms had most impact in the context of orthopaedics. In contrast, the reforms were perceived to have less influence in diabetes and early intervention in mental health. This variation has been largely due to contextual differences in agendas and models of service delivery across the three tracer conditions.

Conclusions The paper responds to the principal feature of realistic evaluation by identifying lessons emerging from the operation of the reforms in order to inform their future development and implementation.