Sadly, yours is a not untypical experience. Profile measures are a dead
duck as far as economic evaluation is concerned *unless* you have the
capacity to aggregate them into a single summary index form.
You have two possibilities for getting out of the 'fix' you're in with
SF-36. You might refer to the work of Denny Fryback who developed a
scoring algorithm for SF-36 based on comparisons with Quality of
Well-Being scale. There is a paper in Medical Decision Making from about
2 years ago. Alternatively, you should consult John Brazier's work on
assigning utility weights to SF-12.
I realise that it is not always possible to influence the choice of
outcome measure. However, it is generally better to know that you have a
viable measure of benefit/outcome *before* collecting data, rather than
after the event. This is particularly true where you are considering CUA
and require a utility-weighted index for quality-adjustment purposes. If
you don't have that - and SF-36 in its native form clearly fails to
deliver such an index - then you might have to rethink the type of
analysis you adopt.
Paul Kind
Fryback, D. et al (1997) Predicting QWB scores from the SF-36. Medical
Decision Making 17(1) 1- 9
David LB Schwappach wrote:
>
> Dear List Members!
> I kindly ask you for help on my question: Do you know of a method to
> "convert" quality-of-life-scores to preference-based utilities? I'm doing a
> cost-utility study and have SF-36 Scores of the patients. Basicly I think
> its a different concept behind the approaches because the SF-36 does not
> incorporate preferences for different health states, but there may be
> something like "preference weights"?
> Any help, advise or suggested reading would be greatly appreciated!
> Can you give me some advise how I could get along with the data I have?
> Maybe you could suggest some reading to me?
> Thank you very much for help for advance!
> Best wishes!
> David
>
> David LB Schwappach
> MSc Econ
> Schlosstr. 34
> 14059 Berlin
> Germany
> e-mail: [log in to unmask]
|