In selecting an instrument users must consider the different types of instrument that are available and how they meet the requirements of the proposed application[1]. Seven major types of instrument are described with examples. They differ in content and also in their intended purpose or application. The classification should not be interpreted too rigidly and is not mutually exclusive.

Disease-specific - eg: Asthma Quality of Life Questionnaire
Population-specific – e.g. Child Health and Illness Profile-Child Edition/CHIP-CE
Dimension-specific - eg: Beck Depression Inventory
Generic - eg: SF-36
Individualised - eg: Patient Generated Index
Summary items - eg: UK General Household Survey questions about long-standing illness
Utility measures - eg: EuroQol, EQ-5D

Disease/condition-specific

These instruments have been developed to measure the patient's perceptions of a specific disease or health problem. Multiple instruments are now available for common health problems.

The Asthma Quality of Life Questionnaire consists of 32 items that produce four dimension scores relating to activity limitations, symptoms, emotional function, and environmental exposure [2].

Advantages The targeted focus of disease-specific instruments can make them clinically relevant. An instrument developed to address a particular disease should be responsive to clinically important changes in health that result from interventions. Disease-specific instruments do not contain any items or health dimensions that are not relevant to the disease. Furthermore, because the instrument has clear relevance to patients with the presenting problem, acceptability is likely to be high.

Disadvantages It is not generally possible to administer disease-specific instruments to samples who do not have the relevant health problem. This means that health status scores cannot be compared with those for the general population, which is a common approach for assessing the impact of a particular disease on health status. It follows that it is not possible to make comparisons across treatments for different diseases, which limits the application of disease-specific instruments in economic evaluation. Finally, the restricted focus of disease-specific instruments may prevent them from detecting side effects or unforseen effects of treatment.

(Back to top)

Population-specific

NB In the literature, the term ‘population-specific’ may be used to describe both disease/condition-specific instruments and those specific to particular demographic groups. Here, we are referring to the latter only.

Population-specific instruments are designed to be appropriate to particular demographic groups, such as children or elderly people.

The Child Health and Illness Profile/CHIP was initially developed for adolescents. A version for children aged between 6 and 11 years, the CHIP-CE, has been developed since. The CHIP-CE consists of 45 items covering the five domains of satisfaction, comfort, resilience, risk avoidance, and achievement [3].

Advantages: The content of population-specific instruments may be more relevant to the group in question - e.g. in the case of young children, the inclusion of items relating to school performance. A specifically tailored format, such as the use of cartoon illustrations to convey instructions rather than text, can make these measures more accessible - enabling individuals who are often not consulted directly to report on their own health and preferences. Population-specific developed instruments may also be sensitive to systematic differences between population groups.

Disadvantages: The disadvantages of using population-specific measures are similar to those of disease-specific measures, for example, ruling out comparisons with the general population, and making it difficult to compare the efficacy of particular treatments across population groups.

(Back to top)

Dimension-specific

Dimension-specific instruments assess one particular aspect of health status. – see Table below which summarises the aspects most commonly identified in the literature as relevant to patient-reported outcomes (see Fitzpatrick et al [1]).

I - Physical function Mobility, dexterity, range of movement, physical activity Activities of daily living: ability to eat, wash, dress

II - Symptoms Pain Nausea Appetite Energy, vitality, fatigue Sleep and rest

III - Global judgements of health

IV - Psychological well-being Psychological illness: anxiety, depression Coping, positive well-being and adjustment, sense of control, self-esteem

V - Social well-being Family and intimate relations Social contact, integration, and social opportunities Leisure activities Sexual activity and satisfaction

VI - Cognitive functioning

Cognition Alertness Concentration Memory Confusion Ability to communicate

VII - Role activities Employment Household management Financial concerns

VIII - Personal constructs Satisfaction with bodily appearance Stigma and stigmatising conditions Life satisfaction Spirituality

IX - Satisfaction with care

The most common type are those that measure psychological well-being.

The Beck Depression Inventory contains 21 items that address symptoms of depression [4]. The instrument was originally developed for use with psychiatric patients but it is increasingly used to assess depression in the physically ill.

Advantages: They provide an assessment of a particular dimension of health that is often more detailed than that provided by disease-specific or generic instruments that attempt to cover broader aspects of health. Many of the instruments, which include measures of physical functioning as well as psychological well-being, have been widely used, so that there is a wide range of data available for comparing and interpreting results.

Disadvantages: Measures of psychological well-being in particular were often developed with the primary measurement objective of discrimination, including the measurement of inter-patient differences for diagnosis or needs assessment. Before such instruments are used in evaluative applications their appropriateness as outcome measures should be examined carefully.

(Back to top)

Generic

Generic instruments are designed to measure very broad aspects of health and are therefore potentially suitable for a wide range of patient groups and the general population.

The SF-36 is one of the most widely used generic instruments [5-7]. It is a 36-item instrument that measures health across eight dimensions of physical functioning, social functioning, role limitations due to physical problems, role limitations due to emotional problems, mental health, vitality, pain, and general health perceptions. Responses to items within each dimension are summed to produce a health profile of eight scores. The dimension scores also form physical component and mental component summary scores [7,8].

Advantages: The main advantage of generic instruments is that they are suitable for use across a broad range of health problems. They can be used for comparisons between treatments for different patient groups to assess comparative effectiveness. They can also be used with healthy populations to generate normative data that can be used to compare different patient groups. Their broad scope means that they have potential to capture the influence of co-morbidity on health, as well as unexpected positive or negative effects of an intervention. This makes them useful for assessing the impact of new health care technologies when the therapeutic effects are uncertain.

Disadvantages: Broad applicability means that some level of detail has to be sacrificed which may limit the relevance of generic instruments when applied to a specific patient population. Generic instruments are potentially less responsive to clinically important changes in health.

(Back to top)

Individualised

Individualised instruments allow respondents to select the content of items and/or rate the importance of individual items.

The Patient Generated Index asks respondents to list the five most important areas of their lives affected by a disease or health problem and then to rate how badly affected they are in each area, and in the rest of their lives [9,10]. They then give a number of 'points' to the areas in which they would most value an improvement. The individual area ratings are weighted by the 'points' given and summed to produce a single index designed to measure the extent to which a patient's actual situation falls short of their hopes and expectations in those areas of life in which they most value an improvement.

Advantages: Individualised instruments address the concerns of the individual patient rather than impose an external standard that may be less relevant. Therefore individualised instruments can have high content validity.

Disadvantages: Individualised instruments have to be administered by interview in order to produce response rates similar to those for standardised instruments. This has implications for the feasibility of individualised instruments when compared to standardised instruments that can be self-administered.

(Back to top)

Summary items

Summary items ask respondents to summarise diverse aspects of their health status using a single item or a very small number of items.

Since 1974 the General Household Survey for England and Wales has used two questions relating to chronic illness and disability: "Do you have any long-standing illness or disability?"and "Does this illness or disability limit your activities in any way?" Transition items are a form of summary item that ask the respondent to assess their current health compared with a specific point in the past, such as their last clinic visit. The SF-36 contains a transition item that asks: "Compared to one year ago, how would you rate your health in general now: excellent, very good, good, fair, poor?"

Advantages: Summary items are brief and make the least demands on respondents' time. Some summary health items have been widely used in large samples of the general population so that there is considerable comparative evidence. Despite their obvious simplicity there is some evidence for the measurement properties of summary items including reliability and validity. Summary items that relate to global health also offer a potential means of exploring apparently contradictory trends in different dimensions of health, for example an improvement in physical function that coincides with a deterioration in psychological well-being.

Disadvantages: The brevity of summary items limits the specific inferences that can be made about particular aspects of health. The number of response categories of summary items are often few and may be too crude to measure small but important changes in health. The general nature of such questions may make them particularly prone to the influence of expectations, transient aspects of mood, and variations between respondents in criteria for answering such questions. Responses to transition items may suffer from recall bias and may be unduly influenced by current health status.

(Back to top)

Utility measures

Utility measures incorporate preferences or values attached to individual health states and express health states as a single index. This type of instrument produces evidence for the overall value of health states to society and can be used in cost-utility analysis.

The EuroQol EQ-5D consists of five items relating to mobility, self-care, main activity, pain/discomfort and anxiety/depression [11,12]. On the basis of their responses to the five items, patients are classified into a health state with a preference weight attached. Preferences for health states are derived from general population surveys using techniques such as the rating scale, standard gamble, and time trade-off. These techniques are sometimes used to obtain direct health state values from patients.

Advantages: Through the incorporation of preferences for health states, utility measures produce a single index. This facilitates comparisons between treatments for different health problems and is useful for economic evaluation including cost-utility analysis.

Disadvantages: Utility measures are usually broad in their focus and are therefore subject to the same criticisms as generic instruments. The difficulty of obtaining health preferences for a large number of states usually means that utility measures are brief. For example, the EuroQol EQ-5D consists of five items that produce 243 health states. There are problems of feasibility with methods of obtaining preferences for health states. Interview-based techniques for eliciting preferences are labour-intensive and time-consuming. Some respondents have difficulty understanding the nature of the experimental tasks they are required to perform.

(Back to top)

References