1. 20 times 80 is enough Ben van Hout 10/4/19
Julius Center for Health Sciences and Primary Care

2. Contents Introductionary remarks Valueing health states
The classic approach
The Bayesian approach
A comparison
How further
Concluding remarks

3. Uncertainty surrounding costs and effects

4. EuroQol

5. EQ-5D Index

6. Valueing health states
Populations from the general public are asked to attain values to health states which are described in terms of scores on different dimensions
EQ-5D
5 dimensions
3 scores per dimension
Not all health states are valued
There is an underlying structure

7. What do we want? To compare the results from different therapies
Using data from RCT’s
Using models
Using valuations from the general public
Medians
Which may be country specific

8. Among the numerous problems
Each country starts its own valuation study withouth learning from the other countries
Utility estimates are hardly ever surrounded with uncertainty margins
Especially when collected alongside trials

9. The MVH-study 3395 respondents 41 health states + 11111 + unconsious
rescaled
15 states per respondent
Mostly about 800 respondents per state
A few with 1300
3333 for all
Inconsistents taken out
39868 valuations

10. The 3074 clean repondents (36369 clean data-points)

11. A typical good health state

12. A typical bad health state

13. A typical health state

14. Mean values + 95% confidence intervals

15. The classic appproach (EQ-5D)

16. Linear model; middle level = 2

17. Linear model; middle level is free

18. Linear Model, middle level free + n3 term

19. Uncertainties surrounding the model estimates

20. Observations + model estimates with 95% confidence limits

21. Let’s go Bayesian
The confidence intervals of my predictions of the average values are sometimes out of the range defeined by the confidence intervals of my observed average values
Wouldn’t it be nice if we would also acknowldege that we are uncertain about our model?
Samer Kharroubi, Tony O’Hagan and John Brazier

22. The Kharroubi approach
The function is unknown and is a random function
The expected values of the function are described by a linear model
Look at all valuations as random variables following a large 243-dimensional multinomial distribution with correlations that decrease with the distance between the states
Respondents may differ by a parameter α.

23. The Kharroubi approach
Succesfull in describing the SF-6D
Unsuccessful in working with 40,000 data-points
So,
A random sample of 38*80 points
Estimate
Predict 3 remaining points
Compare with classical approach

24. Within a few minutes

25. And while waiting the results
What if I don’t use all the data, but just the averages
And play around with a classical Bayesian alternative

26. Parameter estimates all data average estimate se Constant 15.65 0.47
15.38
0.43 B1
-1.49
0.16
-1.58
0.14 B2
-1.33
0.15
-1.28 B3
-0.35
0.17
-0.29 B4
-2.02
-1.92
0.13 B5
-1.29
-1.20 D1
0.88
0.21
0.92
0.20 D2
0.22
0.23
0.12 S3
0.53
0.44 S4 D5
0.56
0.59 N3
-2.80
0.37
-3.03
0.35

27. Estimates based on averages (including 95% confidence intervals)

28. Standard Bayesian

29. Parameter estimates Bayesian Classical estimate se Constant 16.16 0.41
15.38
0.43 B1
-1.62
0.14
-1.58 B2
-1.21
-1.28 B3
-0.59
0.15
-0.29
0.16 B4
-2.08
0.13
-1.92 B5
-1.37
-1.20 D1
0.88
0.18
0.92
0.20 D2
0.12
0.21 S3
0.19
0.44 S4
0.84
0.22 D5
0.60
0.59 N3
-2.31
0.34
-3.03
0.35

30. Aren’t you neglecting something?
Standard Bayesian approach using WinBugs
σ= 0.52 (in comparison to 0.59 following classical approach)
We know the uncertainties surrounding the observed average values
We can include those in WinBugs

31. Partly promising Bayesian

32. Hey, there is Samer Relatively very good fit, but not as good as mine
But much better than mine without the dummies and the n3
My predictions are better

33. A comparison The Bayesian approach is - off course - more intuitive
It seems much more flexible in a natural way
It may take ways more computer-time
It may not handle large data-sets very well
I hope to be more convinced at the end of this week

34. How further
A better inclusion of the uncertainties surrounding the averages
Can’t Samer work with 41 data-points?
Using the first data-set as prior to the next
Designing a next country specific study

35. Concluding remarks Bayesian analysis makes one feel good
Samer for president
I’m almost convinced