2024. 04. 02.
APPLICABILITY OF MEASURING HEALTH RELATED QUALITY OF LIFE IN HEALTH CARE DECISION MAKING
Brno, 18 April 2024 Assoc.Prof. Andräs Inotai
CV Andras Inotai
Affiliation
• Associate Professor (Center for Health Technology Assessment, Semmelweis University)
• 10+ years of teaching experience
• Head of Pharmaceutical Policy Research (Syreon Research Institute)
• 15 years of research experience in pharmaceutical policy, pharmacoeconomics, HTA and patient centric healthcare
Scientific appointment
• President, ISPOR Hungary Chapter/Hungarian Health Economics Association (2014-15) Research track
• 60+ peer-reviewed scientific publications in health economics, pharmaceutical policy, HTA and patient centric healthcare
• Selected list of publications:
https://pubmed.ncbi. nlm.nih.gov/?term=inotai+a&sort=pubdate&size=20
SEMMELWEIS University
Center for Health Technology Assessment
http://semmelweis.hu
SEMMELWEIS Uni
Andräs Inotai:
Did you know that
compared to women men tend to leave shorter, ...but have better health related quality of life?
Areas covered
Paradigm shift in medicine: emerging role of patients and
patient reported outcomes in healthcare
Quality of life, health related quality of life (HRQoL)
Measuring HRQoL, EQ-5D, value set
Health status of the Hungarian population: 2000 vs 2022
Population norms from the perspective of clinicians
Perspective of payers - Quality adjusted life years (QALYs)
1
2024. 04. 02.
Paradigm shift in medicine The emerging role of PROs
• Lower incidence of acute high mortality diseases (infectious diseases) increase life expectancy on developed (and) developing) countries
• Higher prevalence of chronic conditions with disability along with increasing life longevity
• Improving quality of life is becoming increasingly important besides improving life longevity
• Patient reported outcomes (PROs) - information coming directly from patients on their own health status
- Valuing health states based on PROs is relatively new and gaining increasing relevance in the field of Medicine
- Psychometrics: Discipline in psychology to measure PROs - new application in clinical trials
SEMMELWEIS University
Health Outcome Measurement
• Biological/physiological parameters
- (blood cholesterol level mmol/l)
• Observer Reported Outcomes (ObsROs, by non-clinician observer e.g. teacher or caregiver)
- Focus on observable signs of behaviors
• Clinician Reported Outcomes (ClinROs)
- Clinician Global Impression, ECOG Performance Score
• Patient Reported Outcomes (PROs)
- Reported directly: Quality of life
- Reported indirectly: Adherence
Determinants of QoL
2024. 04. 02.
Attributes of health state classification system
Breadth of coverage
physical      emotional      sensory pain cognitive
function       function       function ability
mobility physical activity
self-care role
performance
dressing bathing continence eating
Torrance GW (1986} Journal of Health Economics
Classification of measures
	Profile	Index
Generic measures	NHP, SF-36, SF-12	EQ-5D, SF-6D, HUI2,
		HUI3, QWB
Specific measures	Kidney disease	RAQoL, SGRQ
	questionnaire	
SEMMELWEIS University 0
10
					
	Under eacti fteading, please tick (he ONE box that test cfescrrtes your ;			WeWouldl,«<IIhnDl.hcWV»dorl>ad]r<».heal<l,IsTODAr.                 ^ ^	
	MOBILITY		EQ-5D-3L		u can imagine                                            = 55
	1 iiave no p'cihler-'s n .■;::!!■ "g nhiu*	□		Dm"""""—""""""	
	1 am confined to Bed SELF-CARE 1 tiaue some problems washing or dressing myself lliave some problems with performing my usual activities 1 am unable to perform my usual activities PAIN / DISCOMFORT 1 tiaue no pain or discomfort 1 liave moderate pain or discomfort 1 liave extreme pain o( discomfort ANXIETY ' DEPRESSION	□ □ □ □ □ □ □ □ □	index EQ-VAS		- □    f:
	1 am not amicus or depressed			— *	
	III "III£HT			IBiii ^~	
					
SEMMELWEIS University °                       András Inotai:		Ap Br	EuroQoL © Eu	oQol Group. EQ-5D™ j	a trade mark of the EuroQol Group |
11
EQ-5D-3L index
35=243 possible health states + death + unconsciousness PROs:
- Simple,
- Easy to understand,
- Widely used-international references,
* Comparability among diseases (even with a healthy population, and among different countries
- Several validated translations in different languages to conduct multinational trials CONs:
- Limited sensitivity to detect small changes
12
3
2024. 04. 02.
Value sets
Health states can be described by combination of statements of health-related quality of life measures.
'Value sets' are numerical expressions of how preferred a health state is. Reference points used in value sets:
- 0.0 death (health states considered worse than death have negative weights}
- 1.0 full health
The provision of population-level health-related quality of life estimates (also known as 'population norms') is expected to improve the preciseness of patient-level clinical decision making, health economic and public health studies. However, preference towards these health states is influenced by culture, resulting in differences across populations.
EQ-5D value sets from UK (1997) and Hungary (2020)
Health Status:|ll223| UK
Dimension	Level	Coefficient (UK)	Coefficient (HU)
	Constant	0.081	0.020
Mobility	1. level	0	0
	2. level	0.069	0.022
	3. level	0.314	0.648
Self-care	1. level	0	0
	2. level	0.104	0.051
	3. level	0.214	0.355
Usual activities	1. level	0	0
	2. level	d.d^	0.025
	3. level		
Pain / discomfort	1. level	0	0
	2. level	0.123	0.080
	3. level	0.386	0.338
Anxiety /	1. level	0	0
depression	2. level	0.071	0.078
	3. Level	0.236	0.258
	N3	0.269	
Meas. interval		-0.594 to 1	-0.865 to 1
calculated index value (utility):
1.0 - 0.081-0.036 - 0.123 - 0.236 - 0.269 = 0.255
Health Status: |112231 HU
calculated index value (utility): 1.0-0.020-0.025 -0.080-0.258 =0.617
d et. al. Methods for economic evaluation of health care programs.
Oxford University Press, 1997 z et al.. (2020} Parallel valuation of EQ-5D-3L and EQ-5D-5L by time trade-off in Hungary. VALUE IN HEALTH 23(9}:1235-1245.
14
SEMMELWEIS Uni
HEALTH STATUS OF THE HUNGARIAN POPULATION 2000 vs 2022
Population health survey 2000
• N=5503(age 18+)
• Instrument: EQ-5D-3L
• Value set used to quantify EQ-5D-3L dimensional responses: UK
• Small sample size for age 65+
• Update was necessary
15
16
A
2024. 04. 02.
Characteristics	Total	
N, %	11910	100
Sex		
Women .	6466	54.3
Men	5444	45.7
Age (yearl^^r		
12-17 Lx	578	4.9
18-24	818	6.9
25-34	1456	12.2
35-44	1664	14.0
45-54	1945	16.3
55-64	1892	15.9
65-74	2176	18.3
75-84	1105	9.3
85+	276	2.3
Geographical region {residence)		
Budapest (Central Hungary}	1364	11.5
Pest (Central Hungary}	1005	8.4
Central Transdanubia	1433	12.0
Western Transdanubia	1280	10.8
Southern Transdanubia	1310	11.0
Northern Hungary	1557	13.1
Northern Great Plain	2153	18.1
Southern Great Plain	1808	15.2
Baseline survey data (2022)
IS
Ctd.
Education {highest level completed)*
Less than 8 years in primary school
Primary school
Secondary school
University Occupation**
Employed/self-employed
Unemployed
Student
Other inactive*** Household size
Inotai et al. Population-level n a comparison of survey results from 2022 with 2000. Manuscript under ri
438	3.7
2481	21.0
6447	54.5
2469	20.9
5907	49.6
247	2.1
999	8.4
4023	33.8
723	6.1
2180	18.3
4083	34.3
2349	19.7
1916	16.1
950	8.0
267	2.2
165	1.4
values by EQ-5D-3L in	
Statistical methods
Primary sampling units (PSUs) were settlements, and the secondary sampling units were dwellings in Hungary. The settlements were stratified by county and size.
All household members aged 12 years or older were included in the survey.
Design weights were calculated based on the sampling design. After the data collection, these were calibrated to correct for nonresponse by geographical region to population size, biological sex and age distribution, development category of the settlement, and household size, so that we could provide unbiased estimates on the level of the population.
To include 11910 respondents aged 12 or more, 15058 individuals were contacted in 7578 households, resulting in a response rate of 79%. Of all the EQ.-5D-3L questionnaires, 2.81% were self-administered online, 46.64% by telephone interview, and 50.55% by personal interview, between 1" April and 2nd May 2022. Mean EQ-5D-3L index values with their 95% confidence intervals were estimated for the target population by age and sex using "svy: mean" procedure with analytically derived variance estimator associated with the sample mean.
Inotai et al. Population-level norm values by EQ-5D-3L in Hungary -a comparison of survey results from 2022 with 2000. Manuscript under review
17
18
Dimensional responses
dimension	mobility	self-care	usual activities	pain/ discomf	anxiety/ depr		mobility	self-care	usual activities	pain/ discomf	anxiety/ depr
male 12-15	98.3	98.6	97.8	98.6	99.2.,	P<?male 12-15	99.6	99.8	99.1	97.3	97.4^ 94.9"'
male 16-17	99.2	97.9	97.9	99.2	92.6LJ	[female 16-17	100	100	100	99.2	
male 18-24	97.7	98.3	96.9	96.8	93.5	female 18-24	97.7	97.9	98	95.1	94.7
male 25-34	98.4	99	98.3	94.1	94.2	Ifemale 25-34	98.3	99	97.9	94.8	94.2
male 35-44	96.8	99.8	98.6	92.7	94.2	female 35-44	98.2	99.1	98.2	91.8	93.2
male 45-54	92.9	98.2	94.6	86	90.2	Ifemale 45-54	92.8	98.2	94.9	84.8	87.5
male 55-64	83.1	96.6	88.6	71J n	85.1	Ifemale 55-64	81 n	95.2	87.1	69.3	81.2
male 65-74	63.1 r	90.3	79.9	55 r	80.2	female 65-74	60.7	90.5	76	46.6	70.1
male 75-84	41.5	75.4	56.8	34.3	68.3	Ifemale 75-84	35.81 1	76.3	54.8	26.sJ	58
male 85+	24.3i	• 55.6	36.5	25.9<J	' 51.3	Ifemale 85+	18.7V	48.9	31.8	16.9^	' 50.9
% reporting no problems in each dimension
Inotai et al. Population-level norm values by EQ-5D-3L in Hungary a comparison of survey results from 2022 with 2000. Manuscript under review
Dimensional differences between male and female
	mobility	self-care	usual activities 1	pain/ I discomf 1	anxiety/ depr
12-15J	-1-3	-1.2		1.3	1.8
16-171	-0.8	-2.1		o|	
18-241	0	oa\	-1.1	1.7	-1.2
25-341	0.1	0	0.4	-0.7	0
35-441	-1.4	0.7	0.4	0.9	1
45-54|	0.1	0	-0.3	1.2	2.7
55-64	2.1	1.4	1.5	2.4	3.9
65-74	2.4	-0.2	3.9	8.4	10.1.
75U-	=^ 5.7I	-0.9	2	7.5	10.3
85+	5.6	6.7	4.7	9I	0.4
% reporting no problems in each dimension
Red: female report more problems. Blue: male report more problems
Inotai et al. Population-level norm values by EQ-5D-3L in Hungary -a comparison of survey results from 2022 with 2000. Manuscript under review
19
20
2024. 04. 02.
The impact of value set: whose preference matters?											
	0.9	♦     men 2022 (Hungarianv — A- -men 2022 (UKvalue set			lueset) —		022 (Hungarian value set; 022 (UKvalue set) •J?				
	OJ °-8										
	-o 0.7										
	0.6							X N			
	0.5								& ''X.		
		1B-24	25-34		agecoho	ts 55-54	65-74	75-S4	S3,		
						Inotai et al. Population-level norm values by EQ-5D-3L in Hungary -a comparison of survey results from 2022 with 2000. Manuscript under review					
SEMMELWEIS Unive			András Inotai:		iplicability of m no, 2024 April	ensuring HRQo	in health care	eclslon making			
21
Dimensional changes 2000 vs 2022											
|dimension	mobility self-care usuala			pain/disi anx/depr			mobility	self-care	usualad pain/disc anx/der		
level	L2+L3	L2+L3	L2+L3	L2+L3	L2+L3		L2+L3	L2+L3	L2+L3	L2+L3	L2+L3
2000|men 18-34	2.5	0.9	1.8	14.8	15	men 18-34	0.6	-0.4	-0.4	9.9	9
|men 35-64	17 |	4.8	13.2	35.8	30	men 35-64	8.6	3.1		/> 20.3	20.2
|men 65+	38.9	17.1	30.9	55.9	40.5	men 65+		1	1.5	3.2	15.5
women 18-34	3.2	0.3	2.8	18.2	26.1	women 18-34	1.3	-1.1	0.7	13.1	20.5
|women 35-64	20.7 |	5.3	15.6	48	44.5	women 35-64	11.8	2.9		> 30.5	32.1
|women 65+	49.4	19	35.5	68.4	54.3	women 65+	-2.1	1.1	0.4	5.6	18.6
2022 men 18-34	1.9	1.3	2.2	4.9	6						
men 35-64	8.4	1.7	5.6	15.5	9.8						
|men 65+	45.5	16.1	29.4	52.7	25						
women 18-34	1.9	1.4	2.1	5.1	5.6						
|women 35-64	8.9 |	2.4	6.3	17.5	12.4						
| wo men 65+	51.5	17.9	35.1	62.8	35.7						
% reporting any L2 or L3 problems in each dimension Red: improvement over 22 year, Blue: deterioration over 22 years						Inotai et al. Population-level norm values by EQ-5D-3L in Hungary -a comparison of survey results from 2022 with 2000. Manuscript under review					
SEMM ELWEIS University ° http://5emmelweii.hu		András	-	Appl	ability of me 2024 April 1	asuringHRQoLin healt	hcaredecMo				
22
2024. 04. 02.
Quality of life decrement due to complications of osteoporosis
1	I,	ll - 1
p	= 6	
Si	I i	f i
i!		i1
	--	11
1		1
Error bars represent the 95% confidence intervals * Controlled forage and gender
If
https://pubmed.ncbi.nlm.nih.gov/28895276/
Quality of life decrement due to complications of yy  diabetesjjiellitus ~
11
ü
+
l 2
I i
Cerebrovascular    Pe-ripneml anery
Error bars represent the 95% confidence intervals * Controlled for age and gender
https://pubmed.ncbi.nlm.nih.gov/28895276/
25
26
THE PERSPECTIVE OF DECISION MAKER: QUALITY ADJUSTED LIFE YEARS (QALY)
How to interpret and compare the clinical benefits of health technologies?
50% reduction of epilepsy attacks
50% reduction of tumor size
15% improvement in lung functions
4 points improvement in ADAS-Cog score
20% less patients need symptomatic therapy
2 points improvement in the 10-point pain VAS scale
Which outcome is more important?
27
28
7
2024. 04. 02.
QALY approach
Perfect health
0.5x1
Death 1 2
QALY- (0.5xl)+(0.5x0.25)+(0.5x0.5)+(lx0.75)+(0.5x0.5)=1.875
Treatment outcomes of malignant disorders
Health gain (complex care-^   minimal treatment)
Health capital (complex care]
Life years
https://pubmed.ncbi.nlm.nih.gov/20132213/
SEMMELWEIS University 0
29
30
31
Relevance of QALY
QALY is suitable to aggregate different dimensions of health outcomes (e.g. in the case of complex oncology treatment):
- Long-term life years gain
- Short-term QoL deterioration (due to AEs)
- Long-term QoL improvement
QALY is suitable to estimate the magnitude and sign of aggregate health gain
Set up objective ranking between health technologies through ensuring comparability
- QALY league table
Calculation of QALYs and estimating utilities is standardised Critique of QALY has a wide literature
Full economic evaluation - when is it cost-effective?
New Therapy
Alternative Therapy
32
		SMEW
Impact on health status		
		
		
Impact on healthcare resource use: physical units and costs		
'mpact on health status
QALYhsw - QALYALT ACost
Impact on healthcare	
resource use: physical	.............—
units and costs	
AQALY
incremental cost-effectiveness ratio
8
2024. 04. 02.
33
Take-home messages
• Patient Reported Outcomes earn increased importance in medicine
• HRQoL can be measured by many tools, including EQ-5D-3L instrument
• 'Value sets' are numerical expressions of how preferred a health state (e.g. an EQ-5D-3L state) is.
• Population norm values are used as a reference to estimate the HRQoL decrement of a patient population with different diseases
• QALY is a universal health outcome measure which considers both quality of life (utility) and quantity of life (life longevity)
• QALY is used in cost utility analyses to estimate value for money of health technologies (e.g. pharmaceuticals)
SEMMELWEIS University 0 Andräs Inotai: Applicability of measuring HRQoL in health care decision making
http://5emmelweii.hu Brno, 2024 April 18
34
Implications
• SUKL applies cost-effectiveness analyses for pharmacoeconomic evaluations. Attemps were made to estimate willingness to pay threshold (CZK/QALY)*.
• Interestingly, according to EuroQoL website,
- there are no population norms available for EQ-5D-3L/5L for Czechia
- there is no value set available for EQ-5D-3L/5L for Czechia
• If EQ-5D-3L is used to estimate QALYs, international value sets are available
"Haluska M etal. Cost-Effectiveness Threshold in the Czech Republic: Are ICERS Still Growing? an Updated Analysis. Value Health 2020;23:2S660
35
36
Q