Past CHE Seminars 2007

6 December 2007

Title: Reservation wages, labour market participation and health
: Professor Jennifer Roberts, Department of Economics, University of Sheffield.

Abstract: In this paper, we examine the role of health determining the proportions of the population who are economically active and employed rather than unemployed when economically active. We use data for males from 15 waves of the British Household Panel survey and estimate a model which predicts reservation wages for the unemployed and market wages for the employed. Our results suggest that health is an important determinant of selection, both into economic activity and into employment (versus unemployment) but that, once these participation effects are accounted for, health is not a significant determinant of either the reservation wage or the market wage. Our results have important policy implications since they suggest that poor health is a major cause of economic inactivity and unemployment.

26 November 2007

Title: Using programme budgeting to inform spending decisions in the NHS
Speaker: Dr Peter Brambleby, Director of Public Health, North Yorkshire and York PCT

Abstract: Primary Care Trusts are entrusted with very large sums of public money which they deploy across a range of 'providers' to meet government targets for access to health care but also, ostensibly, to improve the health of their populations.  It should be legitimate to ask a PCT Board: "Where does your money go in the major health areas such as cancer, mental health and circulatory diseases?  What activity does that buy?  What outcomes are generated?  Do you have plans to changing those patterns, and why?  And how does your PCT compare with similar PCTs in all the above?  But most PCTs would struggle to answer.  Some are exploring programme budgeting and marginal analysis as a framework for a more constructive debate. 
Dr Peter Brambleby is one of a growing number of Public Health Directors who is convinced the approach has merit.  He joined North Yorkshire and York PCT in July this year, to build on work he piloted in Norfolk.  He serves on the DH Programme Budget Project Board and related working groups.  This presentation is an opportunity to discuss progress so far, outline the strengths and limitations, explore the practical application of a PBMA approach in the local health economy and seek advice on next steps.

1 November 2007

Title: A review of the epidemiological approach to resource allocation in health care
Speaker: Dr Stephen Morris, Reader, Health Economics Research Group, Brunel University.

Abstract: The need component of the current formulae for allocating resources for hospital services and prescribing in England is based on a utilisation approach. This assumes that expenditure on NHS activity in different areas reflects relative needs and supply conditions and these can be disentangled by multiple regression models to yield an estimate of relative need for NHS expenditure. These assumptions have not gone unchallenged and critics have suggested an alternative epidemiological method.
The epidemiological approach uses direct measures of morbidity to allocate health care resources. It divides the total national budget into sub-programmes based on diagnoses of diseases, computes the proportion of total cases for each sub-programme in each geographical area, and then allocate budgets to geographical areas proportional to their share of total cases.  The main objection to the epidemiological approach has been seen as its very rich data requirements. But there are also methodological limitations of the epidemiological approach. They centre on the assumption of proportionality, which at the small area level requires that the average level of needs for cases is the same in every area. We investigate potential data sources for applying the epidemiological approach in England. We also test the proportionality assumption using data from the 2002-2004 rounds of the Health Survey for England. We find regional variation in disease severity for major diseases, which suggests that health care needs for certain conditions do vary by area, and therefore that the proportionality assumption underpinning the epidemiological approach is unlikely to hold.

27 September 2007

Title: Prioritization and patients’ rights: prioritization practice in the Norwegian Hospital Sector
Speaker: Professor Jan Erik Askildsen, Health Economics Bergen (HEB), Department of Economics, University of Bergen.

Abstract: The Norwegian Patient Rights Act requires that elective patients be prioritized according to the seriousness of their condition and the cost-effectiveness of treatment. Their maximum waiting time depends on their prioritization. 
Because of wide geographical variations in waiting times ownership of hospitals were transferred from local government (19 counties) to 5 regions in 2001.  We compare the period before the reform, 1999-2001, with the period following the reform, 2002-2005. We find that waiting times for prioritization groups have not been equalized across the five regional health enterprises. Within four of the regions differences in waiting times have been reduced. Both across the five regional health enterprises, and within them, the lowest prioritized groups have experienced the relatively largest reduction in waiting times. We discuss possible explanations for the apparent failure of the reform.

12 September 2007  

 Title: Measuring well-being for public policy: preferences or experiences?
Speaker:  Professor Paul Dolan, Chair in Economics, Business School, Imperial College.

Abstract: NICE, along with most health economists, recommends that the weight assigned to health states should reflect our preferences, as expressed through our willingness to exchange extra years of life or the risk of death for improvements in health.  Responses to such preference elicitation based methods for valuing health states, whether from patients or the general public, reflect whatever the respondent thinks about or feels at the time of the assessment, which may not be what they will think about or feel while experiencing that health state.
Policymakers should shift their attention towards valuations derived from more direct measures of the experiences associated with different health states. By using responses to overall life satisfaction questionnaires which also include data on health status, and a wide range of covariates, we can estimate the relative effects of different health states on life satisfaction.
Results suggest that anxiety and depression, even in quite mild forms, have a significant effect on life satisfaction, and limitations in physical functioning generally have much less of an effect.  By contrast preference elicitation studies suggest that mental health and physical health are broadly comparable in terms of their expected impact.

5 July 2007   

Title: Being naughty about NICE? Questioning the methods used to maximize health gains from NHS resources
Speaker:  Professor Stephen Birch, Department of Clinical Epidemiology and Biostatistics, and Centre for Health Economics and Policy Analysis, McMaster University. Ontario.

Abstract: The adoption of economic evaluation guidelines by decision-making bodies in several countries has seen the institutionalization of economics as a basis for decision-making aimed explicitly at getting maximum output from whatever resources are committed to health care.  The National Institute of Clinical Excellence (NICE), perhaps the most widely discussed example of this approach, has been described by Williams (2004) as “the closest anyone has come to fulfilling the economists dream”.  It has been argued that such guidelines are associated less with getting ‘the biggest bang for the bucks’ and more with getting ‘the biggest bucks for a bang’, generally a nightmare for both economists and decision makers.  Following formal reviews of NICE by the House of Commons’ Select Committee (2002) and the WHO (2003),  revised guidelines were developed using a ‘reference case’ as a template for submissions to NICE (National Institute for Clinical Excellence 2004).  Attempts to highlight the implications of underlying assumptions for health care decision making generally have been dismissed by some and heavily criticized by others in their defence of NICE. In this paper we review these criticisms and consider whether the revised NICE guidelines can help turn the economist’s nightmare into Williams’ dream. 

7 June 2007  

Title: Why private sector spatial wage variation matters to the NHS
Speaker:  Professor Bob Elliot, Director, Health Economics Research Unit, University of Aberdeen.

Abstract: The spatial pattern of pay in the private sector of the economy is an important determinant of the distribution of funding within the NHS.  Under Payment by Results the tariff is adjusted by the Market Forces Factor (MFF). The spatial pattern of pay in the private sector also identifies the appropriate patterns of spatial reward for NHS employees. Under Agenda for Change nurses and other professional staff receive High Cost Area Supplements (HCAS). This paper reviews the theoretical arguments underpinning both the MFF and HCAS, and reports an empirical test of the underlying theory. It shows that differences in the patterns of spatial wage premia between the private sector and the NHS affect nursing vacancies, but it also suggests that no similar relationship exists for doctors. It explores the robustness of these results to the underlying assumption that private sector labour markets are in equilibrium, using local unemployment and inactivity rates to measure temporary disequilibrium.

3 May 2007   

Title: Tipping the funding balance: the effect of more long-term (social) care on acute care cost
Speaker:  Dr Julien Forder, Senior Research Fellow, LSE.

16 April 2007

Title: Trading places - touching the void: Estimating domestic values for EQ-5D health states using imported national survey data
Speaker:  Victor Zarate, Centre for Health Economics.

Abstract: Health status measures used to quantify outcomes for economic evaluation must be capable of representing health gain in terms of a single index, usually calibrated in terms of the social preferences of 'the relevant population'. For NICE, these values are considered to be those of the general population of England and Wales and the values obtained through the 1993 MVH project are applied. In many countries where no such social preference weights exist, these MVH weights are (often mistakenly) adopted as the de facto default values for converting EQ-5D into its index form. The survey mechanism for establishing national preference weights for EQ-5D is cumbersome and resource intensive – the MVH study involved 80+ interviewers and 3,395 participants in a protocol that required an hour to complete. The MVH protocol has been used in a modified form in other countries, including Germany, Denmark, Japan, Spain, USA and Argentina. Valuation surveys are currently in train in Thailand, Taiwan and Korea. 

The general problem faced in the majority of countries in which social preferences are required for cost-effectiveness analysis is the absence of a value set based on domestic data sources. This paper describes two alternative strategies for estimating a set of provisional values for EQ-5D health states for use in Chilean economic studies. The first approach is based on a conceptual model of health and leads to a generalisable methodology that might be globally applied. The second approach is data-driven and is based on the 2002 US national valuation survey in which the data obtained from Hispanic have been reworked to provide a proxy for other Latin American countries. Both approaches are novel and are the product of a work programme initiated as part of the Alan Williams Fellowship with the support of the EuroQoL Group.

5 April 2007

Title: Paying for quality: GPs and the quality and outcomes framework.
Speaker:  Professor Hugh Gravelle, Centre for Health Economics, University of York.

Abstract: In April 2004 the NHS introduced the Quality and Outcomes Framework: an extensive system of payment for performance against 146 quality indicators GPs.  The presentation will (a) describe the QOF and its interesting peculiarities; (b) summarise what is know about the effects on the QOF on quality, GP incomes, hours, and job satisfaction; (c) describe the relationship between practice and patient characteristics and QOF performance and the equity implications; (d) examine the incentives for gaming via reporting of prevalence and exceptions, and report tests for gaming; (e) discuss how the QOF might improved.

1 March 2007

Title: An ethical code for decision analysts (including economic evaluators)... and how should we trade-off scientific rigour and practical usefulness?
Speaker:  Professor Jack Dowie, London School of Hygiene & Tropical Medicine.

Abstract: Powerful interests, academic as well as commercial and political, threaten the integrity with which decision analyses of all sorts, including cost-effectiveness analyses and multi-criteria decision analyses, are carried out. Simultaneously they threaten the claims of these techniques to provide alternative decision or decision support technologies. An ethical code is therefore needed to enable professionals to defend key principles of their activity. The draft ‘Montourtier Declaration’ goes well beyond a checklist of best practice requirements to address the wider issues raised when decision analyses of any sort are commissioned or performed. It includes a requirement to address explicitly the trade-off between normativity and practicality in setting ‘prescriptive’ or ‘requisite’ criteria for an analysis – a ‘Value of Analysis Analysis’. We say something about the determination of this trade-off in the context of the meta-decision of deciding how to decide. Much of the ethical code is expressed in terms of a map of the world of judgment and decision making that locates the various Belief, Preference and Decision Technologies on Hammond ’s Cognitive Continuum. The talk begins with a presentation of this map.


Download here

1 February 2007

Title: Using rank and discrete choice data to estimate health state utility values for QALYs
Speaker:  Professor John Brazier, Professor of Health Economics, The University of Sheffield.

Abstract: Aim: There is increasing interest in using ordinal methods for eliciting health state values as an alternative to conventional methods such as standard gamble and time trade-off. Previous work has been hindered by the difficulty of placing results on the full health (=1) and dead (=0) scale required for QALYs.

This paper presents a solution to this problem in the context of using ranking and discrete choice experiment (DCE) data.

Methods: A representative sample of the UK general population (n=308) were interviewed and asked to rank a series of health states derived from the condition specific quality of life instrument from best to worst. Following this exercise the TTO technique was used to elicit values for each health state. Four weeks after the interview, consenting individuals received a postal self-completion questionnaire containing a DCE asking them to make 6 pair-wise comparisons of AQLQ states and 2 of AQLQ states against being dead. An additional 'cold' sample was sent the same questionnaire without the interview beforehand. TTO data were analysed by random effects and OLS models, rank data by rank ordered logit and DCE data by random effect probit. All models assumed an additive effect for the dimension levels, and included a dummy variable for being dead. Rank and DCE results are transformed onto the QALY scale by rescaling the coefficients by dividing the beta coefficients for each dimension level by the coefficient on being dead

Results: There were 308 individuals interviewed (response rate 40%) and 168 of these returned DCE questionnaires. A further 110 from the 'cold' sample returned DCE questionnaires. The TTO model best predicted actual TTO values, followed closely by the rank model (mean absolute errors of 0.051 and 0.065 respectively). The DCE models from the warm and cold samples were very similar and both poorly predicted observed TTO scores. The DCE model predicted higher values for better states and lower values for worse states.

Conclusions: This study has shown how rank and DCE data can be used to generate health state values for calculating QALYs. The rank data generated similar values to TTO, but DCE may generate a wider range of values. Rank and DCE data offer a promising alternative to the more challenging tasks offered by SG and TTO.

Authors: John Brazier, Yaling Yang and Aki Tsuchiya

11 January 2007

Title: Women live longer than men: is this an inequity? And what would gender equity do to it?
:  Dr Aki Tsuchiya, Senior Lecturer in Economics and Health Economics, The University of Sheffield

Abstract: In most modern populations, women live longer than men, and women have a higher level of expected life time QALYs. So, here's an inequality.

The first question addressed in the presentation is whether or not this inequality is inequitable. This was a question pursued in an earlier paper by Tsuchiya and Williams (2004, SSM). One issue mentioned in the paper but not fully explored is that with increasing gender equity, women will pick up male behaviours and attitudes that are not necessarily health promoting, and therefore the inequality itself may diminish over time. This is the second question of the presentation:

What would gender equity do to the existing inequality in health between the sexes. We discuss empirical findings based on Swedish data illustrating how gender equity is associated with different changes in the health of men and women. The presentation is a think piece, and poses more questions than it tries to answer them; the main objective is to get people thinking.

Authors:  Aki Tsuchiya, University of Sheffield, with Anna Månsdotter, Swedish National Institute of Public Health

Who to contact

For more information on these seminars, contact: