Population Health Impact of Disease in Canada (PHI)

Health state preferences

• Overview of preference measurement activities
• Classification system (CLAMES)
• Measurement tools
• Preference measurement
• Scoring function
• FAQs

Overview of preference measurement activities

Preference scores, which indicate the relative preference for a health state compared with full health, help us understand how Canadians view the various aspects of functional health. In practical terms, they allow us to synthesize in a single score the effects of a health state across various attributes of functional health.

Canadian preference scores for various health states were elicited from lay individuals in small groups. Conducting preference exercises in the general population provides a new perspective since other burden of disease studies conducted by the World Health Organization (WHO), the Netherlands , and other countries, used preferences elicited from medical panels.

In panels of about 10 participants, individual Canadians considered how living with the effects of these health states would influence their own lives in terms of usual activities such as work, school, community participation, or family and social roles. Through both individual and group exercises, these panels provided scores for a subset of health states. For each health state, an average score was determined to represent the overall preferences of the Canadian population.

Protocols appropriate to the general public were developed to ensure that participants had a common understanding of each health state and its impact on functional status. All material was thoroughly tested in focus groups and reviewed by experts in the field prior to use. Focus testing was conducted during 2002 and the preference exercises were completed by June 2003.

[return to top]

Classification system (CLAMES)

We developed the Classification and Measurement System of Functional Health (CLAMES) to provide a standardized way to compare various health states. CLAMES provides a standard set of 11 attributes, each with four or five levels ranging from normal to severely limited functioning. The functional limitations associated with the health state are classified by assigning a level for each of the 11 attributes.

Table 1 Classification and Measurement System of Functional Health (CLAMES)

Core Attributes
Pain or Discomfort	1. Generally free of pain and discomfort 2. Mild pain or discomfort 3. Moderate pain or discomfort 4. Severe pain or discomfort
Physical Functioning	1. Generally no limitations in physical functioning 2. Mild limitations in physical functioning 3. Moderate limitations in physical functioning 4. Severe limitations in physical functioning
Emotional State	1. Happy and interested in life 2. Somewhat happy 3. Somewhat unhappy 4. Very unhappy 5. So unhappy that life is not worthwhile
Fatigue	1. Generally no feelings of tiredness, no lack of energy 2. Sometimes feel tired, and have little energy 3. Most of the time feel tired, and have little energy 4. Always feel tired, and have no energy
Memory and Thinking	1. Able to remember most things, think clearly and solve day-to-day problems 2. Able to remember most things but have some difficulty when trying to think and solve day-to-day problems 3. Somewhat forgetful, but able to think clearly and solve day-to-day problems 4. Somewhat forgetful, and have some difficulty when trying to think or solve day-to-day problems 5. Very forgetful, and have great difficulty when trying to think or solve day-to-day problems
Social Relationships	1. No limitations in the capacity to sustain social relationships 2. Mild limitations in the capacity to sustain social relationships 3. Moderate limitations in the capacity to sustain social relationships 4. Severe limitations in the capacity to sustain social relationships 5. No capacity or unable to relate to other people socially
Supplementary Attributes
Anxiety	1. Generally not anxious 2. Mild levels of anxiety experienced occasionally 3. Moderate levels of anxiety experienced regularly 4. Severe levels of anxiety experienced most of the time
Speech	1. Able to be understood completely when speaking with strangers or friends 2. Able to be understood partially when speaking with strangers but able to be understood completely when speaking with people who know you well 3. Able to be understood partially when speaking with strangers and people who know you well 4. Unable to be understood when speaking to other people
Hearing	1. Able to hear what is said in a group conversation, without a hearing aid, with at least three other people 2. Able to hear what is said in a conversation with one other person in a quiet room, with or without a hearing aid, but require a hearing aid to hear what is said in a group conversation with at least three other people 3. Able to hear what is said in a conversation with one other person in a quiet room, with or without a hearing aid, but unable to hear what is said in a group conversation with at least three other people 4. Unable to hear what others say, even with a hearing aid
Vision	1. Able to see well enough, with or without glasses or contact lenses, to read ordinary newsprint and recognize a friend on the other side of the street 2. Unable to see well enough, even with glasses or contact lenses, to recognize a friend on the other side of the street but can see well enough to read ordinary newsprint 3. Unable to see well enough, even with glasses or contact lenses, to read ordinary newsprint but can see well enough to recognize a friend on the other side of the street 4. Unable to see well enough, even with glasses or contact lenses, to read ordinary newsprint or to recognize a friend on the other side of the street
Use of hands and fingers	1. No limitations in the use of hands and fingers 2. Limitations in the use of hands and fingers, but do not require special tools or the help of another person 3. Limitations in the use of hands and fingers, independent with special tools and do not require the help of another person 4. Limitations in the use of hands and fingers, require the help of another person for some tasks 5. Limitations in the use of hands and fingers, require the help of another person for most tasks

[return to top]

The attributes used in CLAMES were adapted from the three most common tools for measuring health status: the Health Utilities Index (HUI 3) ¹ developed by the McMaster Health Utilities Group, the EuroQol Five Dimensions (EQ-5D+)^2,3 and the SF-36 Health Status Questionnaire^4,5 used by Medical Outcomes Trust. Although these instruments had been tested and validated in Canadian studies, none was considered appropriate to define the range of stages of disease and injury examined in the PHI.

CLAMES was developed to cover the spectrum of functioning required for day-to-day living, but with a limited set of attributes so that it was not too complicated for participants to understand in the time available. Seven attributes - Pain or Discomfort, Emotional State, Memory and Thinking, Speech, Vision, Hearing, and Hands and Fingers - were based on HUI 3. Physical Functioning, Fatigue, and Social Relationships were adapted from the SF-36. Anxiety was adapted from EQ-5D+. Table 2 compares the attributes used in each tool.

In some instances the language was simplified (to ensure comprehension by the general population) and for certain attributes, some of the levels were combined. Wording was adapted to keep the language uniform and consistent across attributes.

Table 2 Sources from which CLAMES attributes were adapted

CLAMES	HUI 31	EQ-5D+2,3	SF-364,5
Pain or discomfort	Pain	Pain/discomfort	Bodily pain
Physical functioning	Ambulation	Mobility Self-care Usual activities	Physical functioning Role limitations (physical)
Emotional state	Emotion	Anxiety/depression	Role limitations (emotional)
Fatigue			Energy/vitality
Memory and thinking	Cognition
Social relationships			Social functioning
Anxiety		Anxiety/depression	Mental health
Speech	Speech
Hearing	Hearing
Vision	Vision
Hands and fingers	Dexterity

References

1. Feeny D, Furlong W, Torrance GW, Goldsmith CH, Zhu Z, DePauw S, et al. Multiattribute and Single-Attribute Utility Functions for the Health Utilities Index Mark 3 System. Med Care 2002;40(2):113-28.
2. Rabin R, de Charro F. EQ-5D: a measure of health status from the EuroQoL group. Ann Med 2001; 33: 337-343.
3. Krabbe PF, Stouthard MEA, Essink-Bot ML, Bonsel GJ. The effect of adding a cognitive dimension to the EuroQol multiattribute health-status classification system. J Clin Epidemiol 1999; 52(4): 293-301.
4. Ware JE Jr. SF-36 Health Survey manual and interpretation guide. Boston : The Health Institute, New England Medical Centre; 1993. Ware JE, Kosinski M. How to score the SF-36 physical and mental health summary measures. Boston(MA): The Health Institute; 1994.
5. Ware JE, Kosinski M. How to score the SF-36 physical and mental health summary measures. Boston (MA): The Health Institute; 1994.

Attributes are intended to measure limitations in specific aspects of functioning. This functional status or capacity is related to the individuals' own limitations and not those imposed by their circumstances or surroundings. For instance, Social Relationships assesses a person's internal capacity for developing and maintaining social relationships regardless of opportunities provided by the social environment. For the preference measurement, participants consider how their own situation would be affected by these limitations.

Figure 1 illustrates two description cards used in the preference measurement exercises. Cards were identified by random two-letter codes. Card YD describes mild (chronic) asthma and ML describes severe (chronic) asthma. Note that levels of Pain or Discomfort, Fatigue, and Anxiety are greater and Physical Functioning is more limited for ML than for YD. Core attributes were included for all health states: if there was no limitation, the space was left blank. Supplementary attributes were included only if they were relevant to that health state.

Figure 1 Health state description cards

HEALTH STATE: YD
You have problems with the following:
Pain or Discomfort	Mild pain or discomfort
Physical Functioning
Emotional State
Fatigue
Memory and Thinking
Social Relationships

HEALTH STATE: ML
You have problems with the following:
Pain or Discomfort	Moderate pain or discomfort
Physical Functioning	Mild limitations in physical functioning
Emotional State
Fatigue	Sometimes feel tired, and have little energy
Memory and Thinking
Social Relationships
Anxiety	Mild levels of anxiety experienced occasionally

[return to top]

Measurement tools

The relative preferences of Canadians for various health states were elicited from individual Canadians in small groups using standard techniques. Four methods were tested in focus groups for potential use in the PHI: standard gamble (SG), visual analogue scale (VAS), time trade-off (TTO), and person trade-off (PTO).

With the standard gamble measurement tool, participants were asked to assume that they were in a state of reduced health for the rest of their lives. They were then offered a hypothetical procedure that had a certain probability of restoring them to full health with a corresponding probability of ending their life instantly (for instance, 80% probability of full health and 20% probability of death). They had to choose whether they wished to take the procedure (with the specified probability of success) or to remain in the state of reduced functional capacity for the remainder of their lives. The probabilities were varied until the point of indecision was identified. This was the point at which participants found it most difficult to make a decision. Focus group testing revealed that the standard gamble was the most appropriate approach for measuring preferences in a group setting. It was used as the main preference measurement tool in the PHI.

With the visual analogue scale, participants ranked a series of health states using a thermometer-like scale. The top of the thermometer represents the most desirable health state and the bottom the least desirable health state: the less desirable the participants considered the health state, the closer they placed it to the bottom of the thermometer. Participants of focus groups conducted in the general Canadian population identified that this was a good tool to familiarize them with the concept of ranking health states. It was thus used in the PHI as a training tool before starting standard gamble exercises.

The time trade-off measurement determines the number of years of healthy life an individual would be willing to give up to avoid living in some state of reduced health for 10 years; this is used to calculate a preference score. Focus group testing identified that this exercise was quite difficult to conduct in groups and that there was a lot of confusion about what years were being traded. As a result, the time trade-off method was only used in one group in the PHI, for comparison with other methods.

The person trade-off measurementasks respondents to choose between hypothetical interventions that offer health benefits to groups of individuals in different health states. It identifies the point at which a participant would choose to keep 1,000 individuals in full health alive for one year rather than keep a given number of individuals with less than full health alive for one year. Focus group testing and consultation with experts identified that the person trade-off approach was inappropriate in the Canadian context; the PHI does not use this method.

[return to top]

Preference measurement

Preference scores for health states were elicited in 14 panels of about 10 participants across Canada. Panels were constructed to generally reflect the composition of the Canadian population in terms of age, language, and background.

The visual analogue scale was used as a preliminary training exercise to familiarize participants with the process of rating health states. Standard gamble techniques were then used to elicit each participant's preferences for each health state.

The day-long sessions included group and individual exercises. Each group considered 12 "anchor" states covering a range of health states from full health to dead. The same 12 anchor states were considered by every panel to provide cross-panel comparisons. Participants then individually considered another 14 health states, some of which represented actual diseases and some of which were hypothetical. Preliminary preference scores are an average score given by all individuals considering that health state.

Because there were too many health states to measure directly, a sample of health states was selected for measurement execises. Preference scores for the remaining health states were then determined using mathematical modeling. Health states were chosen to allow two different approaches for developing a model, statistical and decomposed.^1,2

The statistical approach³ uses a regression model, and thus requires direct measurement of a large number of health states. All possible levels of all attributes were measured at least once using a technique called a Latin hypercube sample; this ensures that all possible levels and attributes are adequately represented.

For the second approach, the decomposed approach, we measured a small sample of health states that included "pure states" in which all the attributes are at the best level except one. This is a modification of the approach used to calculate the Health Utilities Index (HUI) multi-attribute utility function (MAUF).^4,5

About 200 health states were evaluated in the individual exercises to provide an appropriate sample of health states for both these methods. Each of these health states was randomly assigned to 6 or more of the 146 participants.

1. Dolan P. Modelling the relation between the description and valuation of health states. In: Murray CJL, Salomon JA, Mathers CD, Lopez AD, editors. Summary Measures of Population Health: Concepts, Ethics, Measurement and Applications. Geneva : World Health Organization; 2002. p. 502-513.
2. Farquhar PH. A survey of multi-attribute utility theory and application. In: Starr MK and Zeleny M, editors. TIMS studies in the management sciences, Vol.6. Multiple criteria decision making. Amsterdam : New-Holland Publishing; 1977. p. 59-89.
3. Brazier J, Roberts J, Deverill M. The estimation of a preference-based measure of health from the SF-36. J Health Economics 2002; 21: 271-92.
4. Torrance GW, Furlong W, Feeny D, and Boyle M. Multi-attribute preference functions: Health Utilities Index. PharmacoEconomics 1995; 7(6):503-520.
5. Le Galès C, Buron C, Costet N, et al. Développement d'un index d'états de santé pondéré par les utilités en population française : le Health Utilities Index. Économie et Prévision 2001;150-1:71-87.

[return to top]

Scoring function

Several functions were fitted to the observed mean preference scores for over 200 health states classified using CLAMES. The most suitable one was a log-linear (multiplicative) function with a scaling adjustment. A simplified version of this function is provided; technical details are available from the authors.

To determine the preference score for a health state, the parameter estimates for all attribute levels are multiplied together. These estimates are displayed in Table 4, and were obtained by fitting the log-linear function to the observed preference scores. In effect, each estimate is a weight for that particular attribute level in determining the preference for a complete health state.

Next, a scaling parameter is applied. This is needed to adjust the score because a multiplicative function can never reach 0, but in the preference elicitation exercises the preference scores for health states range from 0 (death) to 1 (full health). The scaling parameter “stretches” the values downwards towards 0. This adjustment permits scores for health states at 1 to remain at 1, while the lowest estimated value becomes 0. All values in between are accordingly modified, with low values being more affected than high values.

This function can be used to generate preference scores for any of the 10,240,000 health states that are theoretically possible within the CLAMES framework.

Table 3 Example of preference score derived using scoring function

The preference score for breast cancer at diagnosis would be assigned as follows. 1. Based on literature review and expert consultation, a “typical” individual with early stage breast cancer (a very good prognosis) would have the following functional limitations according to the CLAMES framework: Somewhat unhappy (level 3 of Emotional State) Mild limitations in the capacity to sustain social relationships (level 2 of Social Relationships) Moderate levels of anxiety experienced regularly (level 3 of Anxiety) The classification string for this health state is thus 1 1 3 1 1 2    3 1 1 1 1 How to read the classification 2. The appropriate weights for each attribute from Table 4 are multiplied 1* 1 * 0.96 * 1 * 1 * 1 * 0.97 * 1 * 1 * 1 * 1 = 0.93 where 0.96 is the parameter for level 3 of Emotional State (i.e., “somewhat happy”); 1 is the parameter for level 2 of Social Relationships (i.e., “mild limitations in the capacity to sustain social relationships”); and 0.97 is the parameter for level 3 of Anxiety ( “moderate levels of anxiety experienced regularly”); 1 is the parameter for all other attributes, which have no limitations in this health state. 3. The scaling adjustment is applied: (0.93 - 0.115) / 0.885 = 0.922

Note: The value of 0.115 for the scaling parameter is the preference score for the worst possible health state; the value of 0.885 is the complement (1 - 0.115) of the scaling parameter.

Table 4 Parameter estimates for CLAMES function

Attribute level	Pain and Discomfort	Physical Functioning	Emotional State	Fatigue	Memory and Thinking	Social Relationships
1	1	1	1	1	1	1
2	0.98	0.983	1	1	0.985	1
3	0.954	0.949	0.919	0.952	0.985	0.955
4	0.704	0.681	0.719	0.952	0.985	0.915
5	n/a	n/a	0.663	n/a	0.784	0.821

Note: n/a There is no level 5 on this attribute.

Table 4 (continued) Parameter estimates for CLAMES function

Attribute level	Anxiety	Speech	Hearing	Vision	Use of Hands and Fingers
1	1	1	1	1	1
2	0.985	1	0.958	1	0.985
3	0.982	0.956	0.938	0.93	0.985
4	0.833	0.956	0.897	0.884	0.985
5	n/a	n/a	n/a	n/a	0.784

Note: n/a There is no level 5 on this attribute.

FAQs

What do preference scores tell us about health states?
Preference scores, which lie between 0 (dead) and 1 (full health), measure the relative preference of Canadians for each health state. For instance, on average, Canadians would rather live with severe chronic asthma (preference score 0.9) than chronic fatigue syndrome (0.7) or second degree burns (0.4).

How can you compare diseases and injuries that last a short time or are sporadic with those that last a long time?
During health state preference measurement, individuals compare health states as if they lasted the same amount of time. Duration for each health state is incorporated later in the calculation of summary measures so that diseases with longer durations have more impact.

[return to top]