Original Article
Deriving a Preference-Based Single Index from the UK SF-36 Health Survey

https://doi.org/10.1016/S0895-4356(98)00103-6Get rights and content

Abstract

This article presents the results of a study to derive a preference-based single index from the SF-36. The study was an attempt to reconcile a profile health status measure, the SF-36, with the “quality adjusted life years” approach. The study undertook a parsimonious restructuring of the SF-36 using explicit criteria to form the SF-6D health state classification. A sample of multidimensional health states defined by this classification were valued by a convenience sample of health professionals, managers, and patients, who responded to a set of visual analogue scale ratings and standard gamble questions, with highly complete and consistent answers. Statistical models were estimated to predict single index scores for all 9000 health states defined by the new classification. The resultant algorithms can be applied to existing SF-36 data sets and used in the assessment of the cost-effectiveness of health technologies. This preliminary work forms the basis of a larger study currently being undertaken in the UK.

Introduction

Faced with the ever increasing demands for health services, public and private agencies have become interested in the effectiveness and cost-effectiveness of health care interventions. Patient-perceived health has come to be regarded as an important outcome of health care, and hence a measure of effectiveness. An important consequence has been the development and promotion of generic profile measures of health status such as the Short-Form 36 (SF-36) Health Survey [1]. Health status measures are being used extensively in clinical trials, but they have a number of shortcomings for use in economic evaluation [2].

The main approach in health economics has been to value health status in a single unit of measurement known as “quality adjusted life years” (QALYs) 3, 4, 5 or “well years” [6], which combine increased life expectancy and improvements in health status. Evaluations of health care technologies using this measure present results in terms of incremental cost per QALY for each intervention [5]. This approach uses an index to “quality adjust” survival, thus a person expected to survive 10 years at a quality of 0.8 has eight QALYs. The scale is designed so that 0 equals death and 1 is equivalent to full health.1 The earliest work in this field was undertaken in North America by Torrance 3, 7, 8 and Bush, Kaplan, and colleagues 6, 9, 10, and in the UK, by Williams and colleagues 4, 11. However, QALYs are not being widely used in clinical trials. One reason for this has been evidence of the poor descriptive validity and sensitivity of some QALY instruments compared with existing health status instruments 12, 13, 14, 15.

This article presents a study which attempts to reconcile a profile health status measure, the SF-36, with the QALY approach by deriving a single index measure based on people’s preferences. It begins with a more detailed discussion of the problems in using the SF-36 in economic evaluation in its current form. This is followed by a discussion of alternative approaches to deriving a single index, including the one adopted here, which is partly based on earlier work by Torrance and colleagues at McMaster 7, 16. The methods and results of the three key research tasks of the study are presented: the reduction of the SF-36 into a classification amenable to valuation, the valuation of a sub-sample of health states defined by this classification, and the use of statistical methods to estimate an algorithm for valuing all health states defined by the classification. This is followed by a critical discussion of the use of the results of this study and proposed further work.

Section snippets

Background

The SF-36 Health Survey is a standardized questionnaire used to assess patient health across eight dimensions of health [1]. It consists of items or questions on each health dimension. Responses to the items are combined into dimension scores mainly using simple summation. The SF-36 physical functioning dimension, for example, has 10 items to which the patient can make one of three responses: “limited a lot,” “limited a little,” or “not limited at all.” These responses are coded 1, 2, and 3,

Methods

There are a number of methods for deriving a single index from a profile instrument such as the SF-36. The simplest is to combine the dimension scores or item responses into a single index using an assumed set of weights 28, 29. Such an arbitrary procedure would be unlikely to reflect preferences for different aspects of health. Furthermore, for use in cost-utility analysis the index would have to be combined with survival to form QALYs. O’Brien and colleagues contemplated a similar problem in

background of respondents

One hundred and seventy-three people were asked to participate in the valuation study, and 165 agreed (a response rate of 95%), 55 of whom had been recruited in hospital outpatient clinics. A range was achieved in terms of age (16–79 years) and gender (51% male). The majority were in non-manual occupations or were students, and half reported limiting long-standing illnesses, disabilities, or infirmity.

completion rates

There were 1745 visual analogue scale valuations completed by respondents out of a total of

Discussion

The results of this study can be applied to any existing SF-36 data set. The first task is to map the SF-36 item responses onto the SF-6D using a simple computer program available from the author. The second is to value the resulting SF-6D health states using the coefficients of the “consistent” models. The predicted VAS and SG values of health state 224244, for example, can be calculated as follows:

Estimates of the 23 dummy variables representing decrements from full health did not support the

References (42)

  • J Ware et al.

    Overview of the SF-36 Health Survey and the International Quality of Life Assessment (IQOLA) Project

    J Clin Epidemiol

    (1998)
  • C Donaldson et al.

    Should QALYs be programme-specific?

    J Health Econ

    (1988)
  • J.E Brazier

    The SF-36 Health Survey and its use in pharmaco-economic evaluation

    Pharmaco-economics

    (1995)
  • G.W Torrance

    Social preferences for health statesAn empirical evaluation of three measurement techniques

    Socoiecon Plan Sci

    (1976)
  • A Williams

    Economics of coronary artery bypass after grafting

    Br Med J

    (1985)
  • M.F Drummond et al.

    Methods for the Economic Evaluation of Health Care Programmes

    (1987)
  • R.M Kaplan et al.

    Health statusTypes of validity and the index of well-being

    Health Serv Res

    (1976)
  • G.W Torrance

    Multi-attribute utility theory as a method of measuring social preferences for health states in long-term care

  • G.W Torrance

    Measurement of health state utilities for economic appraisalA review

    J Health Econ

    (1986)
  • R.M Kaplan et al.

    Health-related quality of life measurement for evaluation of research and policy analysis

    Health Psychol

    (1982)
  • R.M Kaplan et al.

    The General Health Policy ModelAn Integrated Approach. Quality of Life Assessments in Clinical Trials

    (1990)
  • P Kind et al.

    Valuation of quality of lifeSome psychometric evidence

  • R Carr-Hill et al.

    Current practice in obtaining the ‘Q’ in QALYsA cautionary note

    Br Med J

    (1991)
  • J Brazier et al.

    Testing the validity of the Euroqol and comparing it with the SF-36 health survey questionnaire

    Qual Life Res

    (1993)
  • R Harper et al.

    A comparison of outcome measures for patients with chronic obstructive pulmonary disease in an outpatient setting

    Thorax

    (1997)
  • D Feeny et al.

    A comprehensive multi-attribute system for classifying the health status of survivors of childhood cancer

    J Clin Oncol

    (1992)
  • C.A McHorney et al.

    The MOS 36-item Short-Form Health Survey (SF-36)II. Psychometric and clinical tests of validity in measuring physical and mental health constructs

    Med Care

    (1993)
  • J.E Brazier et al.

    Validating the SF-36 health survey questionnaireNew outcome measure for primary care

    Br Med J

    (1992)
  • C.A McHorney et al.

    The MOS 36-item Short Form Health Survey (SF-36)III. Tests of data quality, assumptions and reliability across diverse patient groups

    Med Care

    (1994)
  • A.M Garrett et al.

    The SF-36 health survey questionnaireAn outcome measure suitable for routine use within the NHS

    Br Med J

    (1993)
  • Cited by (582)

    View all citing articles on Scopus
    View full text