1000minds supports research into the measurement and valuation of health-related quality of life.
Health-related quality of life (HRQoL) is the perceived quality of a person’s life with respect to the physical and emotional aspects of their health, such as their experience of pain, disability, depression, etc.
HRQoL is measured and valued for countries’ populations as a whole for the purpose of calculating Quality-Adjusted Life Years (QALYs) when performing Cost-Utility Analysis (CUA) and for creating Patient-Reported Outcome Measures (PROMs). For example, CUA is used to support health technology prioritization and PROMs are used to assess the performance of health care providers.
A veritable global industry has grown up around creating systems for describing people’s HRQoL. Well-known health descriptive systems include:
- HUI (Health Utilities Index)
- 15D (15 Dimensions)
- AQoL (Assessment of Quality of Life)
- PROMIS (Patient-Reported Outcomes Measurement Information System)
- SF-6D (Short Form, 6 Dimensions)
- EQ-5D (EuroQoL, 5 Dimensions)
- EQ-HWB (EuroQoL, Health and Wellbeing)
For example, the EQ-5D, which is the most widely used descriptive system in the world, represents HRQoL on five dimensions: mobility, self-care, usual activities, pain/discomfort and anxiety/depression. The most recent version of the EQ-5D, the EQ-5D-5L, has five levels on each dimension – e.g. no, slight, moderate, severe and extreme problems – and hence the EQ-5D-5L represents 3125 (55) health states.
A 1000minds tool for creating HRQoL value sets
A value set consists of HRQoL index values for all health states representable by the particular descriptive system used – e.g. 3125 values for the EQ-5D-5L’s 3125 states (as referred to above). Health state values are anchored at one for full health and zero for dead, with negative values for states considered to be worse than dead.
A specialised 1000minds tool is available for creating both personal and social value sets – i.e. for each individual participant and also a population overall – for any of the health descriptive systems above.
The tool, which includes extensive checks of the quality of each participant’s data, implements the PAPRIKA method (a type of adaptive discrete choice experiment, DCE) and a binary search algorithm to identify any health states worse than dead (News »).
Potentially 1000s of people can be surveyed to obtain a representative sample of a country’s population. Compared to other approaches, the 1000minds tool significantly reduces the cost and time involved in creating and analyzing value sets.
The tool could also support CUA and PROMs at the individual patient level, incorporating the patient’s preferences into treatment decisions in ‘real time’. For example, the tool could be available on computer tablets in doctor waiting rooms or as a mobile app for patients to quickly create their own personal value sets.
- A social value set for the EQ-5D-5L for New Zealand has been created and is available for use now.
- A social value set for the SF-6D for NZ is currently being developed and is expected to be available by the end of 2021.
How significant are correlations between EQ-5D value sets?
As explained in the article below, high correlation coefficients for EQ-5D value sets derived from different samples – across countries and/or using different valuation techniques – are conventionally interpreted as evidence that the respective samples have similar health-related quality of life (HRQoL) preferences. However, EQ-5D-3L and EQ-5D-5L value sets contain many inherent rankings of health state values by design.
By calculating correlation coefficients for value sets created from random data, we demonstrate that ‘high’ coefficients are artifacts of these inherent rankings – e.g. median Pearson’s r = 0.783 for the EQ-5D-3L and 0.850 for the EQ-5D-5L instead of zero. Therefore, high correlation coefficients do not necessarily constitute evidence of meaningful associations in terms of similar HRQoL preferences.
Based on simulations, we calculate significance levels and find that many high coefficients reported in the literature are not statistically significant. These ‘high’ but insignificant correlations are in fact spurious.
Here is our tool to lookup the statistical significance of correlation coefficients for EQ-5D-3L and EQ-5D-5L value sets:
If you’d like to test significance for a different model, we may be able to use the significance calculator we created for the above article to generate significance tables for you.
F Ombler, M Albert & P Hansen (2018), “How significant are ‘high’ correlations between EQ-5D value sets?”, Medical Decision Making 38, 635-45.