Properties of the health state descriptive system
Feasibility and general applicability: These properties are usually measured by response and completion rates. For 15D they are high, comparable to those of similar kinds of preference-based generic instruments, e.g., EQ-5D.
Reliability: Here the questions is about repeatability of measurements with a minimum of random error = test-retest repeatability. For 15D the Bland-Altman (1986) repeatability coefficients are 92-100 % depending on dimension. In this respect there are no significant differences between different instruments.
Validity: Here the question is about the degree of confidence that can be placed in the inferences drawn from the scores of a measure. Therefore validation is a continuous process by which experiences are gained from how the instrument performs is different situations and how credible results it produces. Unfortunately there is no gold standard against which to judge different instruments, but one has to resort to indirect evidence say in terms of content validity and construct validity.
Content validity: The 15D is obviously the most comprehensive of the preference-based generic instruments in sampling dimensions for the construct of HRQoL. It covers most of the “domains of health” emphasised in the WHO International Classification of Functioning, Disability and Health (ICF), see Chatterji S et al. The conceptual basis for measuring and reporting on health. Global Programme on Evidence for Health Policy Discussion Paper No 45, WHO 2002.
- Construct validity: There are clear and substantive convergent and discriminant evidence of construct validity provided by multitrait-multimethod matrix correlations and extreme group comparisons (see Sintonen H. The 15D-measure of health-related quality of life. I. Reliability, validity and sensitivity of its health state descriptive system. National Centre for Health Program Evaluation, Working Paper 41, Melbourne 1994 http://business.monash.edu/__data/assets/pdf_file/0009/391374/wp41-1.pdf; Richardson J, Iezzi A, Khan MA, Chen G, Maxwell A. Measuring the Sensitivity and Construct Validity of 6 Utility Instruments in 7 Disease Areas. Med Decis Making. 2016 Feb;36(2):147-59.
Sensitivity: Sensitivity has two aspects: Discriminatory power and responsiveness to change.
Discriminatory power = ability to distinguish between individuals and groups in different health states cross-sectionally
Theoretically 15D has high discriminatory power, since it defines an enormous number of different health states. In this respect the 15D has more discriminatory power that any other of the preference-based generic instruments.
- Empirical results have confirmed the better discriminatory power (less ceiling and floor effect, more “well-behaved” distributions) of the 15D in comparison to other preference-based generic instruments, see e.g., Sintonen H. The 15D-measure of health-related quality of life. I. Reliability, validity and sensitivity of its health state descriptive system. National Centre for Health Program Evaluation, Working Paper 41, Melbourne 1994 http://business.monash.edu/__data/assets/pdf_file/0009/391374/wp41-1.pdf;
Hawthorne G, Richardson J, Day NA. A comparison of the Assessment of Quality of Life (AQoL) with four other generic utility instruments. Ann Med 2001; 33: 358-370
As a further example of the superior discriminatory power (less ceiling effect) of the 15D in comparison to EQ-5D in a representative Finnish population sample aged 30+ years 47 % was classified as being in ”full health” by EQ-5D, but only 15 % by the 15D. In the same data EQ-5D classified 6-39 % of people with one of 29 chronic diseases/conditions as being in ”full health” contrary to 1-11 % by the 15D (Saarni et al. The impact of 29 chronic conditions on health-related quality of life: a general population survey in Finland using 15D and EQ-5D. Qual Life Res 2006; 15(8):1403-1414).
Another example is about 600 patients prior to coronary artery bypass grafting (CABG) or percutaneous transluminal coronary angioplasty (PTCA), of whom EQ-5D classified 15 % as being in ”full health”, 15D none (Kattainen 2004).
Responsiveness to change = ability to detect changes in individuals or groups over time. This is an important property in follow-up studies.
Theoretically the 15D is highly responsive, since it defines an enormous number of different health states to move from and to. Responsiveness on different dimensions can be seen from the differences in profiles.
Empirical results have confirmed the better responsiveness to change, (when assessed from the changes in the total scores). Examples:
6 months after CABG or PTCA the HRQoL of 79 % of patients had improved according to the 15D, according to EQ-5D 51 % (Kattainen et al. 2005).
Larger effect sizes and higher responsiveness statistics values than EQ-5D (based on UK TTO valuations) in the treatment chronic obstructive pulmonary disease (COPD) (Stavem K. Reliability, validity and responsiveness of two multiattribute utility measures in patients with chronic obstructive pulmonary disease. Qual Life Res 1999; 8: 45-54).
Larger effect sizes and higher responsiveness statistics values than EQ-5D, HUI2, HUI3 and QWB-SA (similar to SF-6D) in the rehabilitation of musculoskeletal, cardiovascular and psychosomatic disorders (Moock J, Kohlmann T. Comparing preference-based quality-of-life measures: results from rehabilitation patients with musculoskeletal, cardiovascular, or psychosomatic disorders. Qual Life Res 2008 Apr;17(3):485-495).