Introduction To Patient-reported Outcome Measures
This document was uploaded by user and they confirmed that they have the permission to share it. If you are author or own the copyright of this book, please report to us by using this DMCA report form. Report DMCA
Overview
Download & View Introduction To Patient-reported Outcome Measures as PDF for free.
More details
- Words: 1,637
- Pages: 33
Patient Reported Outcome Measures Dr Justin Whatling 2009
Clinical Director Routine Health Outcomes Ltd measuring health Visiting Professor Centre for Health Informatics & Multiprofessional Education © Routine Health Outcomes Ltd
Agenda • Introduction to PROMs • Case studies ▫ BUPA Case Study ▫ howRU Case Study
• Health utility exercise
3
Measuring health • Healthcare here for only 2 reasons ▫ Live longer ▫ Improve quality of life
• Yet we only routinely measure mortality Adding Years to Life and Life to Years
© Routine Health Outcomes Ltd
Renewed interest High Quality Care for All If quality is to be at the heart of everything we do, it must be understood from the perspective of the patient. Just as important [as clinical measures] is the effectiveness of care from the patient’s own perspective which will be measured through PROMs Darzi. High Quality Health For All: NHS Next Stage Review Final Report. London: Department of Health, 2008
The ultimate measure by which to judge the quality of a medical effort is whether it helps patients (and their families) as they see it. Donald Berwick, BMJ 1997
Linking revenue to outcomes shows that DH is serious Payment to hospitals for services is being linked to patient-reported experiences and outcomes as one way of driving improved quality and patient-focus across the NHS HM Government. Working Together – Public Services on Your Side. London: HMSO, March 2009
Outcomes
Adapted from FDA/ Harmonisation meeting 16/02/01. Rockville MD
What are PROMs • PROMs are questionnaire tools that measure patient’s health status ▫ Measuring the same questionnaire over time enables comparison and change in health status to be assessed ▫ Change in health status can be assessed against patient progress or health interventions received • Main types ▫ Health Related Quality of Life, HRQL patient’s evaluation of the impact of a health condition and its treatment on relevant aspects of life
▫ Health Utility a cardinal measure of the preference for, or desirability of, a specific level of health status or specific health outcome
• Quality of life is subjective….. ▫
“Given its inherently subjective nature, consensus was quickly reached that quality of life ratings should, whenever possible, be elicited directly from patients themselves.” Neil Aaronson 1996
From research to practice • Outcomes research “seeks to understand the end results of particular healthcare practices and interventions”. (Agency for Health Care Quality and Research, US Dept of Health and Human Science)
• Outcomes management is “the enhancement of physiologic and psychosocial patient outcomes through development and implementation of exemplary health practices and services driven by outcomes assessment.” (Wojner AW, President Health Outcomes Institute)
© Routine Health Outcomes Ltd
Something for everyone
• Research
• Patients and families
• Clinicians
▫ ▫ ▫ ▫
What is the likely outcome and recovery path? How am I doing? What should I spend my personal health budget on? Does my clinician really understand how I am feeling?
• Commissioners ▫ ▫ ▫
Is this population group better than that one? Is there any difference in outcome between care providers? How can we allocate resources to achieve better value for money?
▫ ▫
What is the best treatment for this condition? What is the cost-effectiveness of an intervention
▫ ▫ ▫ ▫
Has this patient’s condition changed for better or worse? Am I missing something that is really important for the patient? How do my results compare with others? Support revalidation
• Managers ▫ ▫ ▫
How do our results compare with others? What are we doing that provides little or no benefit? How can we evaluate cost-saving innovation to check that quality does not suffer?
Outcomes Transparency
The Virtuous Circle
PROM tools • Dimension specific ▫ Single dimension such as function or pain ▫ e.g. Hospital anxiety and depression scale (HADS), McGill Pain Questionnaire • Disease/ Condition specific ▫ Single condition ▫ Thousands of specialised instruments ▫ e.g. Arthritis Impact Measurement Scale (AIMS), Parkinson's Disease Questionnaire (PDQ-39) • Generic ▫ All conditions, all care settings ▫ Relatively few instruments ▫ e.g. Short Form-36, Sickness Impact Profile, EQ-5D
© Routine Health Outcomes Ltd
History of Generic PROMs • First Generation (1968-1988) ▫ Measure health service output ▫ Clinician as the rater ▫ e.g. Rosser, Williams, RAND • Second Generation (1988-2008) ▫ Clinical trials of pharma products, research ▫ Patient as the rater ▫ e.g. SF-36, SF-12, EQ-5D, HUI • Third Generation (2008-)
▫ Routine use at point of care ▫ e.g. howRU
Measuring routine outcomes
Beyond discrete surgical interventions
• Routine health outcomes capture required • Scalable solutions required
Approach • Objective – ‘fit for routine use’ ▫ Quick and simple ▫ Appropriate technology ▫ Deliver value to care professionals and consumers • New ▫ Learn from mistakes of others ▫ Limited evidence base so far • Generic ▫ Multiple conditions & care settings health care, social care, wellbeing, work
What matters to us as patients Pain or discomfort Feeling low or worried
Limited in what you can do Dependent on others © Routine Health Outcomes Ltd
howRU
© Routine Health Outcomes Ltd
Length in words
© Routine Health Outcomes Ltd
Early Validation Study • Patients with 21 long term conditions • Telephone survey • Summer 2008 (beta version phone script) • 2,908 subjects • Picker Institute • Comparison with SF-12
© Routine Health Outcomes Ltd
howRU Phone Script (beta version) We’d now like to ask about how you are feeling today and how much you can do. How are you today…. do you have any of the following: Symptoms, such as pain? none; slight; quite a lot; extreme. Feeling low or worried? none; slight; quite a lot; extreme. Limited in what you can do? none; slight; quite a lot; extreme. Dependent on others? none; slight; quite a lot; extreme.
Correlation 0.99
Correlation 0.97
Psychometric properties • Good internal structure, howRU ▫ ▫ ▫
Item pair correlations are significant at the 0.01 level (2-tailed) Average inter-item correlation is 0.50, at top end of the expected range Suggests that the howRU dimensions are not independent and measure different aspects of the same underlying continuum
• Support for a single summary score, howRU ▫ ▫ ▫
The correlations between each item and the aggregate howRU score have a mean of 0.61, which are substantial and suggests that a single summary score is appropriate to use Principal factor analysis of the howRU items using maximum likelihood extraction suggests that a single latent dimension underlies these items Supported by Cronbach’s alpha, which is 0.8 (expect 0.7-0.9)
• Good correlation with SF-12 ▫ ▫ ▫ ▫
Physical howRU dimensions are more highly correlated with PCS (mean 0.66), than with MCS (mean 0.32) Mental howRU dimension is more highly correlated with MCS (0.59) than with PCS-12 (0.33) The correlation between the aggregate howRU score with the aggregate SF-12 score (sum of PCS-12 and MCS-12) is 0.81 Exploratory principal components analysis: the component loadings suggest that howRU physical dimensions together with PCS-12 load substantially onto one component, and that howRU mental dimension and MCS-12 load onto the second (orthogonal) component.
howRU Properties • Universal ▫ All conditions, all care settings • Takes almost no time ▫ Seconds not minutes • Improves patient care ▫ helps patients and clinicians in routine use • Integrate with IT systems ▫ Record linkage and risk adjustment • Easy to understand results ▫ Simplicity
Health Economics • Comparing different allocations ▫ What should we spend our money on Wheel chairs? Cancer screening?
▫ Can measure cost of inputs but not outputs – can’t put an economic value on healthcare So we use cost-effective analysis Measure costs and outcomes Cost per outcome
Quality Adjusted Life Years (QALY) • Outcomes must be comparable ▫ Disease specific outcomes are incompatible and only allow comparisons within a disease area ▫ Generic outcome are compatible and allow comparisons between disease areas ▫ Best outcome we currently have is QALYs
• Multiply life years with a quality of life index ▫ Perfect health 1.0 to death 0.0
• Then can establish the Cost per QALY ▫ This is the metric used by NICE ▫ Stated range £20,000 - £30,000 per QALY
QALY Example • A new wheelchair for elderly (iBOT) ▫ ▫ ▫ ▫ ▫ ▫
Increases quality of life = 0.1 10 years benefit Extra costs: $3,000 per life year QALY = Y x V(Q) = 10 x 0.1 = 1 QALY Costs are 10 x $3,000 = $30,000 Cost/QALY = $30,000/QALY
30
Cost per QALY league table Intervention
$ / QALY
GM-CSF in elderly with leukemia
235,958
EPO in dialysis patients
139,623
Lung transplantation
100,957
End stage renal disease management
53,513
Heart transplantation
46,775
Didronel in osteoporosis
32,047
PTA with Stent
17,889
Breast cancer screening
5,147
Viagra
5,097
Treatment of congenital anorectal malformations
2,778
How to measure utility • Ask the patient or members of the public? ▫ Quality of life is subjective….. “Given its inherently subjective nature, consensus was quickly reached that quality of life ratings should, whenever possible, be elicited directly from patients themselves.” Neil Aaronson 1996
▫ VAS scores wheelchair patients 8.0, view of health controls 8.3 (Scan J Rehab Med 1985) ▫ Is health a consumer market where consumer values count?
How to measure utility • Indirect utility assessment ▫ HUI, EQ-5D, SF-6D leading but with different strong and weak points ▫ More states ≠ better sensitivity, large number of health states requires statistical techniques
• Direct utility assessment ▫ Standard Gamble, Time Trade Off, Person Trade Off, Visual Analogue Scale
Finally • Patient Reported Outcome Measures have multiple practical uses, all driving to help improve the quality and cost-effectiveness of service provision to patients • Along with clinical outcomes and patient experience, they are an important piece of the healthcare jigsaw and as such are fundamental to health system transformation
www.routinehealthoutcomes.com [email protected]
Clinical Director Routine Health Outcomes Ltd measuring health Visiting Professor Centre for Health Informatics & Multiprofessional Education © Routine Health Outcomes Ltd
Agenda • Introduction to PROMs • Case studies ▫ BUPA Case Study ▫ howRU Case Study
• Health utility exercise
3
Measuring health • Healthcare here for only 2 reasons ▫ Live longer ▫ Improve quality of life
• Yet we only routinely measure mortality Adding Years to Life and Life to Years
© Routine Health Outcomes Ltd
Renewed interest High Quality Care for All If quality is to be at the heart of everything we do, it must be understood from the perspective of the patient. Just as important [as clinical measures] is the effectiveness of care from the patient’s own perspective which will be measured through PROMs Darzi. High Quality Health For All: NHS Next Stage Review Final Report. London: Department of Health, 2008
The ultimate measure by which to judge the quality of a medical effort is whether it helps patients (and their families) as they see it. Donald Berwick, BMJ 1997
Linking revenue to outcomes shows that DH is serious Payment to hospitals for services is being linked to patient-reported experiences and outcomes as one way of driving improved quality and patient-focus across the NHS HM Government. Working Together – Public Services on Your Side. London: HMSO, March 2009
Outcomes
Adapted from FDA/ Harmonisation meeting 16/02/01. Rockville MD
What are PROMs • PROMs are questionnaire tools that measure patient’s health status ▫ Measuring the same questionnaire over time enables comparison and change in health status to be assessed ▫ Change in health status can be assessed against patient progress or health interventions received • Main types ▫ Health Related Quality of Life, HRQL patient’s evaluation of the impact of a health condition and its treatment on relevant aspects of life
▫ Health Utility a cardinal measure of the preference for, or desirability of, a specific level of health status or specific health outcome
• Quality of life is subjective….. ▫
“Given its inherently subjective nature, consensus was quickly reached that quality of life ratings should, whenever possible, be elicited directly from patients themselves.” Neil Aaronson 1996
From research to practice • Outcomes research “seeks to understand the end results of particular healthcare practices and interventions”. (Agency for Health Care Quality and Research, US Dept of Health and Human Science)
• Outcomes management is “the enhancement of physiologic and psychosocial patient outcomes through development and implementation of exemplary health practices and services driven by outcomes assessment.” (Wojner AW, President Health Outcomes Institute)
© Routine Health Outcomes Ltd
Something for everyone
• Research
• Patients and families
• Clinicians
▫ ▫ ▫ ▫
What is the likely outcome and recovery path? How am I doing? What should I spend my personal health budget on? Does my clinician really understand how I am feeling?
• Commissioners ▫ ▫ ▫
Is this population group better than that one? Is there any difference in outcome between care providers? How can we allocate resources to achieve better value for money?
▫ ▫
What is the best treatment for this condition? What is the cost-effectiveness of an intervention
▫ ▫ ▫ ▫
Has this patient’s condition changed for better or worse? Am I missing something that is really important for the patient? How do my results compare with others? Support revalidation
• Managers ▫ ▫ ▫
How do our results compare with others? What are we doing that provides little or no benefit? How can we evaluate cost-saving innovation to check that quality does not suffer?
Outcomes Transparency
The Virtuous Circle
PROM tools • Dimension specific ▫ Single dimension such as function or pain ▫ e.g. Hospital anxiety and depression scale (HADS), McGill Pain Questionnaire • Disease/ Condition specific ▫ Single condition ▫ Thousands of specialised instruments ▫ e.g. Arthritis Impact Measurement Scale (AIMS), Parkinson's Disease Questionnaire (PDQ-39) • Generic ▫ All conditions, all care settings ▫ Relatively few instruments ▫ e.g. Short Form-36, Sickness Impact Profile, EQ-5D
© Routine Health Outcomes Ltd
History of Generic PROMs • First Generation (1968-1988) ▫ Measure health service output ▫ Clinician as the rater ▫ e.g. Rosser, Williams, RAND • Second Generation (1988-2008) ▫ Clinical trials of pharma products, research ▫ Patient as the rater ▫ e.g. SF-36, SF-12, EQ-5D, HUI • Third Generation (2008-)
▫ Routine use at point of care ▫ e.g. howRU
Measuring routine outcomes
Beyond discrete surgical interventions
• Routine health outcomes capture required • Scalable solutions required
Approach • Objective – ‘fit for routine use’ ▫ Quick and simple ▫ Appropriate technology ▫ Deliver value to care professionals and consumers • New ▫ Learn from mistakes of others ▫ Limited evidence base so far • Generic ▫ Multiple conditions & care settings health care, social care, wellbeing, work
What matters to us as patients Pain or discomfort Feeling low or worried
Limited in what you can do Dependent on others © Routine Health Outcomes Ltd
howRU
© Routine Health Outcomes Ltd
Length in words
© Routine Health Outcomes Ltd
Early Validation Study • Patients with 21 long term conditions • Telephone survey • Summer 2008 (beta version phone script) • 2,908 subjects • Picker Institute • Comparison with SF-12
© Routine Health Outcomes Ltd
howRU Phone Script (beta version) We’d now like to ask about how you are feeling today and how much you can do. How are you today…. do you have any of the following: Symptoms, such as pain? none; slight; quite a lot; extreme. Feeling low or worried? none; slight; quite a lot; extreme. Limited in what you can do? none; slight; quite a lot; extreme. Dependent on others? none; slight; quite a lot; extreme.
Correlation 0.99
Correlation 0.97
Psychometric properties • Good internal structure, howRU ▫ ▫ ▫
Item pair correlations are significant at the 0.01 level (2-tailed) Average inter-item correlation is 0.50, at top end of the expected range Suggests that the howRU dimensions are not independent and measure different aspects of the same underlying continuum
• Support for a single summary score, howRU ▫ ▫ ▫
The correlations between each item and the aggregate howRU score have a mean of 0.61, which are substantial and suggests that a single summary score is appropriate to use Principal factor analysis of the howRU items using maximum likelihood extraction suggests that a single latent dimension underlies these items Supported by Cronbach’s alpha, which is 0.8 (expect 0.7-0.9)
• Good correlation with SF-12 ▫ ▫ ▫ ▫
Physical howRU dimensions are more highly correlated with PCS (mean 0.66), than with MCS (mean 0.32) Mental howRU dimension is more highly correlated with MCS (0.59) than with PCS-12 (0.33) The correlation between the aggregate howRU score with the aggregate SF-12 score (sum of PCS-12 and MCS-12) is 0.81 Exploratory principal components analysis: the component loadings suggest that howRU physical dimensions together with PCS-12 load substantially onto one component, and that howRU mental dimension and MCS-12 load onto the second (orthogonal) component.
howRU Properties • Universal ▫ All conditions, all care settings • Takes almost no time ▫ Seconds not minutes • Improves patient care ▫ helps patients and clinicians in routine use • Integrate with IT systems ▫ Record linkage and risk adjustment • Easy to understand results ▫ Simplicity
Health Economics • Comparing different allocations ▫ What should we spend our money on Wheel chairs? Cancer screening?
▫ Can measure cost of inputs but not outputs – can’t put an economic value on healthcare So we use cost-effective analysis Measure costs and outcomes Cost per outcome
Quality Adjusted Life Years (QALY) • Outcomes must be comparable ▫ Disease specific outcomes are incompatible and only allow comparisons within a disease area ▫ Generic outcome are compatible and allow comparisons between disease areas ▫ Best outcome we currently have is QALYs
• Multiply life years with a quality of life index ▫ Perfect health 1.0 to death 0.0
• Then can establish the Cost per QALY ▫ This is the metric used by NICE ▫ Stated range £20,000 - £30,000 per QALY
QALY Example • A new wheelchair for elderly (iBOT) ▫ ▫ ▫ ▫ ▫ ▫
Increases quality of life = 0.1 10 years benefit Extra costs: $3,000 per life year QALY = Y x V(Q) = 10 x 0.1 = 1 QALY Costs are 10 x $3,000 = $30,000 Cost/QALY = $30,000/QALY
30
Cost per QALY league table Intervention
$ / QALY
GM-CSF in elderly with leukemia
235,958
EPO in dialysis patients
139,623
Lung transplantation
100,957
End stage renal disease management
53,513
Heart transplantation
46,775
Didronel in osteoporosis
32,047
PTA with Stent
17,889
Breast cancer screening
5,147
Viagra
5,097
Treatment of congenital anorectal malformations
2,778
How to measure utility • Ask the patient or members of the public? ▫ Quality of life is subjective….. “Given its inherently subjective nature, consensus was quickly reached that quality of life ratings should, whenever possible, be elicited directly from patients themselves.” Neil Aaronson 1996
▫ VAS scores wheelchair patients 8.0, view of health controls 8.3 (Scan J Rehab Med 1985) ▫ Is health a consumer market where consumer values count?
How to measure utility • Indirect utility assessment ▫ HUI, EQ-5D, SF-6D leading but with different strong and weak points ▫ More states ≠ better sensitivity, large number of health states requires statistical techniques
• Direct utility assessment ▫ Standard Gamble, Time Trade Off, Person Trade Off, Visual Analogue Scale
Finally • Patient Reported Outcome Measures have multiple practical uses, all driving to help improve the quality and cost-effectiveness of service provision to patients • Along with clinical outcomes and patient experience, they are an important piece of the healthcare jigsaw and as such are fundamental to health system transformation
www.routinehealthoutcomes.com [email protected]
Related Documents
Outcome
June 2020 8
Reliability Of Outcome Measures And Data Sources
December 2019 12
Measures
November 2019 34
Outcome 4
June 2020 5
