Sample for utility

Asked: 12 Apr 2018 | 1227

In terms of number of respondents required for utility data, is there any standardized formula to calculate? And whether it is justifiable to calculate using proportion estimation? In case of the limited sample (e.g. rare disease, low survival rate) and do not meet the expected sample (as calculated) that may cause wide standard error and affect QALY, could the sample still be used for the analysis? In this case, adding the number of samples would be extremely difficult. If we borrowed utility from other studies (other countries), any adjustment to be considered? and how to measure the transferability? Many thanks.

Utility measure

Expert Replies:

Alec Morton

Professor | United Kingdom | Replied: 15 Apr 2018 at 17:43

It's hard to reply without knowing more details of what specific condition you are looking at and what purpose you have in mind for the utility scores. However, I would say that if you are concerned that your sample may not be large enough to give a sufficiently precise estimate, it would certaintly be insightful to compare with other countries. You could always check to see whether the sample characteristics, like age, disease stage, comorbidity etc from the other countries match with the sample and the population from your own country. If the populations are indistinguishable or almost indistinguishable and you can obtain the raw data for the other country studies, you could construct a pooled utility estimate for a group of countries.