Quality of Life Measurement

Robert M. Kaplan University of California, San Diego Department of Family and Preventive Medicine NIH RCT Summer Training Course August 5, 2002

Where This is Going
Profile Approach
SF-36

Utility Approach
QWB

Preference Assessment

Health-Related Quality of Life is:
What the person can DO (functioning)
Self-care Role Social

How the person FEELs (well-being)
Emotional well-being Pain Energy

HRQOL is Multidimensional HRQOL Physical Mental Social .

Types of HRQOL Measures Profile Generic Targeted Preference-based .

RAND-36 Scales (Items) Physical functioning (10 items) Role limitations/physical (4 items) Role limitations/emotional (3 items) Social functioning (2 items) Emotional well-being (5 items) Energy/fatigue (4 items) Pain (2 items) General health perceptions (5 items) .

Physical Functioning Item Does your health now limit you in bathing or dressing yourself? Yes. not limited at all . limited a little No. limited a lot Yes.

A little of the time.Emotional Well-Being Item How much of the time during the past 4 weeks have you been a very nervous person? None of the time. Most of the time. A good bit of the time. All of the time . Some of the time.

Transform raw average or sum to 0100 possible range (linear transformation) (raw score – minimum)* 100/(max – min) .Scoring RAND-36 Scales Average or sum all items in the same scale.

HRQOL of HIV Infected Adults Emotional Physical MS ESRD Diabetes Depression Prostate disease GERD Epilepsy General Pop CDC C CDC B CDC A 0 20 40 60 80 100 Hays. et al. American Journal of Medicine . (2000).

W. 52. 11-19. R. Wells.. . (1995). K. Archives of General Psychiatry. Rogers.Course of Emotional Well-being Over 2years for Patients in the MOS General Medical Sector 81 79 77 75 73 71 69 67 65 63 61 59 57 55 Hypertensio n Diabetes X X Subthreshol d Depression Major Depression Baseline 2-Years Hays.. & Spritzer.B. C.D. Sherbourne... K.D. Functioning and well-being outcomes of patients with depression compared to chronic medical illnesses.

. et al. and Stewart. A. D. Marshall. Psychological Assessment.L. G. 441449. Hays. Fouryear cross-lagged associations between physical and mental health in the Medical Outcomes Study. The structure of self-reported health in chronic disease patients.D. (1994). • . 2. R. N..Two Underlying RAND-36 Dimensions • Hays. 22-30. (1990). R. 62. Journal of Clinical Psychology.

Indicators of Physical Health Physical Health Physic al functio n Role functio nphysica l Pain General Health .

Indicators of Mental Health Mental Health Emotion al WellBeing Role function emotion al Energy Social function .

energy/fatigue Intercorrelation = 0. R. general health perceptions Mental Health Composite Emotional well-being.91 Hays. D. S. role—physical. pain. Embury. (1998).66. & Chen. H. reliability >= 0. . RAND-36 Health Status Inventory.. social functioning. San Antonio: The Psychological Corporation.RAND-36 Summary Scores Physical Health Composite Physical functioning. role—emotional.

Range of Treatment Impacts on PCS 12 10 8 4 2 0 Treatment Outcomes Duodenal Ulcer Medication Shoulder Surgery mpact on SF6 36 PCS Heart Value Replacement Total Hip Replacement .

Range of Treatment Impacts on MCS 12 10 8 Stayed the same Low back pain therapy Hip replacement Ulcer maintenance Recovery from Depression mpact on SF6 36 MCS 4 2 0 Treatment Outcomes .

149). Pharmacoeconomics MCID for SF36 is “typically in the range of 3 to 5 points” (p. (1999).Samsa et al. .28 ES .09->0.

but MCS was only 0. Role—physical. . and social functioning 1. energy/fatigue 1. (1998. Med Care). Med Care).20 SD lower. 194 MS patients Emotional well-being was 0.Caution in Using SF-36 PCS and MCS Simon et al.7 SD lower. Physical functioning. role-emotional 0. 536 primary care patients initiating antidepressant tx. but PCS did not change! Nortvedt et al.49 SDs.0 SD lower.0 SD lower than general US population. pain. and general health perceptions improved significantly and by 0.28 to 0.3 SD lower. (2000.

Limitation of RAND-36: Is New Treatment (X) Better Than Standard Care (O)? 100 90 80 70 60 50 40 30 20 10 0 0 X 0 0 X X X 0 Physical Functioning X>0 Pain 0>X Emotional Well-being 0 >X Social Functioning X=0 .

0012588) . 49.00001931) .0014261) + (RP x GH * 0.0011709) .00000705) + (PF x BP * 0.59196 + (PF * 0.00001140) + (BP x EWB * 0.9% of the observed QWB variance. Prediction of QWB from SF-36 56.5% on on cross-validation QWB~ = 0.(EWB * 0.Fryback et al.(BP * 0.

Summary of RAND-36 Generic profile measure Single integrated score Preference-based measure Estimate of preference-based measure .

QWB Approach .

1995) . does the patient benefit from the health care they are given? (From Kind.Outcomes Measurement Does the health care you give. affect patient health status? How do you know? How do you distinguish between + and effects on health status? OVERALL.

Traditional • Life Expectancy • Infant Mortality • Disability Days .

0 • Dead 0.0 .Survival Analysis • Alive 1.

Problem with Survival Analysis • Tennis player • Man in coma 1.0 1.0 .

Purpose of Quality Adjusted Survival Analysis • To summarize life expectancy with adjustments for quality of life .

Quality of Well-being Scale Currently two versions Interviewer Self-Report Takes about 10 minuets Automated scoring. low cost About 200 published papers describe use .

QWB Components Functional Scales Mobility (MOB ) Physical Activity (PAC) Social Activity (SAC) Symptom/Problem Complexes (CPX) .

Purpose of Quality Adjusted Survival Analysis • To summarize life expectancy with adjustments for quality of life .

Mobility Scale No limitations in travel • Did not drive or use public transportation • .

Physical Activity Scale Walked without physical problems • Walked with limitations • Moved own wheelchair without help • Confined to bed or chair • .

Social Activity Scale • • • • • Did work. but not work. school. school or housework. school or housework and other activities Did work. school. but limited in other activities Limited in amount or kind of work. or housework Had help with self care . or housework Performed self-care.

remembering. or thinking clearly pain in back or neck sick or upset stomach coughing wheezingof breath spells of feeling upset.Symptoms or Problems (selected) • • • • • • • • • coma trouble learning. depressed or of crying overweight runny nose problems with sexual interest or performance .

Quality-Adjusted Life Year Combines morbidity and mortality into a single index Represents life expectancy with adjustments for quality of life Is defined as a year of life free of all disabilities and symptoms .

605 For each year in this state.Example Case: 68 year old COPD patient Description • • • • • Shortness of breath Drove Car In Bed or Chair for Most of Day Performed No Major Role Activity.395 well years 1- .605 = . the patient loses . but did perform self-care Weight • • Peer Rating equals .

95 0. Preliminary 1.85 0.90 Sinusitis Diabetes Emphysema Average QWB 0.80 0.65 0.55 0.Sinus Disease and Diabetes in the General Population Source: Erickson. 1980 NHIS.00 0.70 0.75 0.50 15-44 45-64 65+ Age .60 0.

QWB by Level of Cognitive Impairment in Alzheimer’s 800 775 750 725 700 675 650 625 600 575 550 525 500 475 450 425 400 Control Mild Moderate Severe Group .

60 1 2 3 4 Serum Beta2 Microglobulin Quartile .80 QWB 0.QWB and Serum Beta 2 Microglobulin in HIV QWB by Serum Beta2 Microglobulin 0.75 0.85 0.65 0.70 0.

50 0.75 0.QWB and Neurological Evaluation in HIV QWB by neurologist rating of central impairment 0.65 0.60 0.40 QWB QWB 1 2 3 4 5 Rating of Central Impairment .80 0.55 0.70 0.45 0.

1992) 44 42 FSI SCore 40 38 36 34 32 30 1 2 3 QWB Quintile 4 5 .Atrial Fibrillation (Ganiats et al.

75 0.50 Dead.55 0.80 0.70 QWB 0.65 0.60 0.QWB and Survival in HIV 0. N=46 Alive. N=466 Status .

Estimating treatment effects Quality Adjusted Life Expectancy for Non-smokers QWB Quality Adjusted Life Expectancy for Smokers Effect of Smoking in QALYs Years .

68 0.74 0.72 0.64 0.6 0. 1994) 0.58 0.56 0.66 QWB 0.7 0.54 Follow-up .Patients Undergoing Sinus Surgery Vs Control (Hodgson.62 0.

80 0.50 0.30 0.60 QWB 0.40 0.20 0.10 0.00 Pre Evaluation Post . 1990) 0.70 0.QWB Before and After Ciprofloxacin Treatment for Exacerbations of CF (Orenstein et al.

0 Weeks .Issues in Child Health: Chronic Episodic Chronic Asthma 1.0 0.

and Being to long .QWB has been criticized for Excluding mental health Excluding sensory function Excluding social health Excluding disease specific information.

Is Mental Health Excluded from the QWB?

QWB by SAPS Patient Groups and Controls
0.8 0.7 0.6 0.5 0.7051 Overall F = 28.49, df 3/118, p<.001 Linearity F = 81.6, df 1/118, p<.001 0.6605 0.5486 0.5196

QWB

0.4 0.3 0.2 0.1 0

Controls

Mild
(0 - 2)

Moderate
(3 - 6)

Severe
(7 - 14)

Number of Positive Psychiatric Symptoms Figure 1

QWB by Hamilton Depression (from Rubin et al 1994)
0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0

QWB Score

Control

0-9

10 to 16

>=17

Hamilton Depression Groups

QWB-SA Mental Health trouble sleeping felling upset and blue excessive worry feeling no control feeling lonely frustration hangover change in sexual interest or performance memory loss thoughts images mediation loss of appetite .

58 Beck-Hamilton .Correlations with Depression QWB-Hamilton .70 QWB-Beck . Beck=.30 .69 Using older QWB weights r HamD=.33.

64 0.49 0.34 0.69 0.79 0.89 0.54 0.74 0.59 0.39 0.44 0.94 0.84 0. N=301) .Frequency 0 10 20 30 40 50 60 70 80 90 0.29 0.99 1 QWB-SA Distribution (Andresen 1998.

Summary QWB and SF-36 have some common roots Correlations between QWB and some SF-36 components are substantial QWB now can be self-administered QWB can be used to estimate QALYs for policy analysis Several theoretical and technical issues must be resolved in future studies .

Utility Assessment Issues .

0 0.The difficult task of development of population utility weight for a health condition 0 1.0 Individuals .72 Population utility weight 0 Population distribution 1.0 0 1.

Preference and Utility Assessment Standard Gamble Time Trade-off Rating Scales .

p Choice B Prob. 1989) .Choice A State A for Rest of Life Prob. p-1 Healthy for Rest of Life Death Example of Standard Gamble (Torrance & Feeny.

Category Scale Perfect Health Visual Analog Death Perfect Health Death .

Approaches for describing health states Brief Outline Age: 40-65 yr. Social activity: did not perform but performed self-care activities Symptoms/problems: sick or upset stomach. I have lost 35 pounds in the last 6 months. I am tired and weak and walk with the aid of a walker. or diarrhea Narrative text I am in the age rage of 40-64 years I live alone and am confined to my home. I require assistance to get into and out of the bathtub. Social contact with my family and friends is infrequent. Multimedia What is an appropriate strength of stimulus to form preferences? . old employee housekeeper Mobility: in house Physical activity: walks but has limitations. vomiting. I am able to only eat small amounts of food at present and I vomit occasionally.

Perceived state attributes Utility Risk attitude Time preference Health preferences Elicitation procedure Utility measurements Health State Description Cognition Emotions and prejudices Numeracy Random error Logical error Cross method inconsistency Anchoring on single values .

Potential Sources of Variability In Preference Measurements Descriptions of states Scaling methods Measurement or assignment procedures Health Status and Social factors .

Validity No absolute standard or threshold Valid aggregate utilities can only arise from valid measurements within individuals Validity of measurements within an individual measurable by Ability to discriminate among states within the protocol Internal consistency of individual responses .

Medical Decision Making 1993) rated their quality of life either 0.9 or 0.0: 15% Results of the Beaver Dam study (Fryback et al.0: 44% . quality of life some other value:18% assigned value of 1.Discrete distributions of utilities make can assessment of reliability difficult.5: 23% rated their quality of life at 1.

1999 .015 10 20 30 40 50 60 70 SG Rating 80 90 100 Satisfy Fail Lenert and Treadwell.62 90 80 70 60 50 40 30 20 10 0 0 r=0. Medical Decision Making.Procedural Invariance and the Organization Of Preferences within an Individual 100 90 80 70 VAS Rating VAS Rating 60 50 40 30 20 10 0 0 10 20 30 40 50 60 70 SG Rating 80 90 100 100 r=0.

Comparison between ever and never in wheelchair or walker for 31 items: Data from Oregon Health Services Commission 100 Wheelchair versus Not in Chair r=.97 80 Never in Wheelchair or Walker 60 40 20 0 0 20 40 60 80 100 Ever in Wheelchair or Walker .

8 Utility 0.1.0 0.9 Impaired in 2 Dimensions Impaired in 1 Dimension Minimal Impairment Standard Gamble0.6 States with impairments in 2 dimensions of health States with impairments in 1 dimension of health States with Minimal impairments .7 0.

46 QWB Rating 0.42 In bed Wheelchair Limited Walking Physical Activity .45 0.Additive Independence and Interval Scaling 0.44 0.47 In House In Hospital 0.43 0.

000 per QLAY $1K 0.05 0.1 Annualized gain in utility .000 per QLAY Annualized marginal cost $50.01 0.000 per QLAY $10.Potential minimally important differences $10K $100.

Summary Competing methods (RS. Each capture different aspects of preferences. SG) are really Complementary methods all have some evidentiary basis.TTO. . The focus of research needs to be: understanding of the implications of choice of a particular scaling method effects of changes in procedures for elicitation. There is no gold standard for measurement of preferences.

Sign up to vote on this title
UsefulNot useful