Quality of Life Measurement Among Persons with Chronic Mental Illness: A Critique of Measures and Methods

Next | Previous | Table of Contents

1. Introduction

1.1 Preamble

Over the last ten years, consumer based appraisal of therapeutic effectiveness has been used with increasing frequency as a central component of health care evaluation and clinical trials. This growth in consumer voice is in part due to a growing appreciation of psychosocial determinants of health status and, in part, to a philosophic shift towards empowering individuals to play an active role in promoting their own health (Wilson, 1996). Consequently, decision-makers are beginning to seek consumer feedback and participation during the planning, delivery, and evaluation of health care services. Typically, such evaluative feedback from consumers, significant others, and their caregivers, are sought in an attempt to evaluate the (perceived) impact of treatment on patients' symptomatology, functional status, welfare, safety, satisfaction, and life quality (Awad & Hogan, 1994; Goodinson & Singleton, 1989; Revicki, Turner, Brown & Martindale, 1992; Turner, 1994).

Contributing to this trend towards consumer based evaluation is increasing fiscal pressure on management to lower program expenditures and, at the same time, to maintain or improve patient satisfaction with services. The recent interest among Canadian mental health professionals in client-based outcome indicators to both monitor population health status and evaluate program effectiveness has had a longer history on the other side of the Atlantic, where services are routinely evaluated in terms of their broad impact on the life quality and functioning of patients and where evaluation methods have been developed to derive service impact estimates by weighing the impact of treatment on patients' quality of life against the costs of providing treatment. In order to facilitate comparisons and to account for the duration and severity of the illness, cost-benefit estimates are standardized across various disease conditions. One widely used example is the Quality Adjusted Life Years (QALY) approach (Rabin, Rosser & Butler, 1993).

In step with these trends, Canadian and US evaluators of mental health services continue to develop appropriate service evaluation models and methods with which to foster efficiency within, and satisfaction with, their service delivery systems (Campbell, 1996, Jenkins, 1990). In their National Strategy for Research on Schizophrenia, the Canadian Alliance for Research on Schizophrenia (CAROS, 1994) cited the identification and evaluation of comprehensive, integrated systems of treatment and rehabilitation, and the development and evaluation of programs for multidiagnosed mentally ill persons as key research objectives in the area of service delivery. In particular, the employment of client outcomes to determine the relationships of program elements to effectiveness was emphasized as an essential component of mental health services research.

Priority was also placed on the measurement of client outcomes, in relation to quality of life, at the 2nd National Round Table on Mental Health, co-sponsored by the Canadian Psychiatric Association and the National Network on Mental Health in 1995.

In concert with the above-noted CAROS and Round Table initiatives, Health Canada has continued to actively support the development of a knowledge base for outcome measurement in the area of chronic and severe mental disorders (Health Canada, 1994a, 1994b, 1995, 1996a).

In 1996, the Federal/Provincial/Territorial Advisory Network on Mental Health identified the need to examine appropriate models and indicators of mental health reform in the interests of exploring linkages between models of care, and program and system delivery, on the one hand, and successful outcomes of mental health services interventions, on the other hand (Health Canada, 1996b).

This interest in health care evaluation activities at the national, provincial and local levels has attracted many converts to the field. Newcomers who are insufficiently trained in research and evaluation often find, however, that the selection of measurement tools and evaluation designs poses some significant challenges. All too often, instrument identification and selection become a chance event, more influenced by word of mouth and apparent face-validity (see discussion on instrument validity in Chapter 3 of this paper) than by a sound matching of population, purpose, instrument and method. Further complicating evaluation initiatives, mental health patients' functional limitations (e.g., in relation to exercising judgement, comprehension and affective state) often compound the difficulties routinely associated with self-report measurement, including the tendency of respondents to answer in an expected way (i.e., social desirability pull) and the response bias associated with changes in affective state (Atkinson, Zibin & Chuang, 1996).

Perhaps one of the most difficult tasks facing anyone beginning to explore the bio-psychosocial effects of illness on the life quality of patients is settling on a workable definition for the construct of Quality of Life (QoL). Arriving at such an operational definition is particularly troublesome since quality of life is, by nature, a complex and multidimensional construct. Individuals' subjective experience of life quality is impacted by an intricate interplay of objective circumstances which change over time. Their experiences are mediated by their unique views, values and appraisals (Corten, Mercier & Pelc, 1994). For example, QoL may be influenced by experiences which change over the lifespan and across various social roles as children, students, family members, employees, patients and church members, to name a few. Moreover, the quality of experience within each of these areas of life is affected by internal states of physical functioning, emotional well-being and personal values. Like external conditions in the physical and social environments, these internal states and value systems may also change and interact in complex, unpredictable ways.

It is also important to keep in mind that formal mental health services are only one of the important influences that affect the lives of persons with mental disorders. Other fundamental components include the role of family and friends, other services (e.g., social services) and supports, and such things as housing, education, income and work (Canadian Mental Health Association, 1993).

Costain, Hewison & Howes (1993) elegantly acknowledge the conceptual difficulties inherent in QoL measurement and suggest that evaluators need to explicitly state the conceptual framework on which their observations rest. Two conceptual perspectives on quality of life are broadly acknowledged by QoL researchers; one is based on a functional view of life quality and the other is based on an illness-experience model of life quality. The functional model focuses on various "objective" aspects of physical and mental functioning which are impacted by illness. Not surprisingly, this model is favoured by evaluators and researchers with an interest in medical management and clinical trials. The illness-experience model, on the other hand, tends to focus on more "subjective" or experiential aspects of the disease process. Those favouring this model place primary importance on the subjective experiences of distress and disability associated with illness. This latter group tends to be more psychosocially focused and para-medically oriented.

Given the conceptual and methodological difficulties already mentioned, the primary objective of this work is to provide an informational resource and discussion paper for those considering the use of QoL assessment in their clinical practice, evaluation activities or research projects. This document reviews current QoL instruments, measurement issues, and evaluation methods, all of which are important if QoL assessment is to be implemented in a meaningful way. Twenty eight (28) QoL instruments are reviewed and discussed with respect to the domains of life they cover, the psychometric properties of their scales, and their use as reported within the recent literature. Various methodological and application issues are addressed, including: the validity of measurement methodologies, the pro's and con's of various evaluation designs, and the possible role of QoL measurement in mental health planning and policy formulation.

1.2 Organization of the Report

This document is divided into four parts, each addressing topics and issues which are pertinent when implementing Quality of Life evaluation within mental health programs. In Chapter 2, 28 instruments are reviewed as to the domains of QoL they address, the methods by which they are administered, and the types of data they collect. Chapter 3 first reviews the fundamentals of instrument reliability and validity (also known as "psychometric properties") and, based on evidence presented in recent literature, provides psychometric ratings for each instrument. Chapter 4 addresses some of the central issues which QoL evaluation raises at both the micro and macro level of service delivery planning and policy formation. Finally, Chapter 5 presents a more detailed psychometric description as well as sourcing information for the QoL instruments reviewed in Chapters 2 & 3. References cited in Chapters 1 to 5 appear after Chapter 5. Other references that may be of interest are listed in Appendix C.

Chapters 2 and 3 contain a set of detailed tables which are provided as an aid to instrument selection. Table 1 classifies the QoL instruments according to the major bio-psychosocial-spiritual domains of life which they address. Table 2 presents an overview of instruments describing the method and time requirements for administration. Table 3 is provided to help clarify the conceptual framework utilized by each instrument as to whether they utilize "objective" or "subjective" types of measurement scales. Finally, Table 4 presents our evaluation of the reliability and validity of each QoL instrument.

While the current document is not intended to offer quick solutions or to substitute for a sound grasp of issues and methods in the field, it is intended to centralize and organize some of the materials necessary to make good measurement and/or project management decisions. We recognize, however, the limitations of this work since literature searches may be incomplete in some areas and their content influenced by date of publication, journal membership and availability and language of publication.

1.3 Literature Search & Review Strategies

An extensive literature search of both the MedLine and the Psychological Abstracts CD-Rom databases between January 1991 and January 1996 forms the basis of the current review (see Appendix A for the search specifications). All abstracts from primary references which addressed QoL measurement, research methods and health policy issues within mental health were selected and hard copies obtained, either locally or through inter-library loans.

Each of the 386 abstracts was read, duplicates were eliminated and 137 articles were selected for examination. These articles were read twice, each time for different content. A first reading sought to identify exhaustively all QoL instruments used with the target populations over the last five years. Some instruments were included which either showed promise but were as yet untested with a mental health population, or utilized some unique measurement methodology. For instruments which were not developed within the last five years, the original studies were examined. These articles were used to evaluate the psychometric properties of the selected instruments (referred to in Chapters 3 and 5). The psychometric ratings of the instruments were based on a consideration of the reliability and validity of scales (i.e., internal consistency, test-retest and alternate forms, and face, construct, content, convergent and predictive validity) as they pertain to use with various chronic mental health populations. This rating is only intended as a rough sketch of psychometric properties and, for reasons which will be discussed in Chapter 3, should only be considered as one piece of the puzzle when selecting instruments for clinical evaluation activities.

A second reading of the initial search abstracts sought to target articles which contained a general review of issues related to QoL measurement in health care including current health policy and planning issues (referred to in Chapter 4).

Next | Previous | Table of Contents