Measuring quality-adjusted life-years when health fluctuates

Recurrent fluctuations in health states can occur due to long-term conditions with episodic symptoms or through side effects of cycles of treatment. Fluctuations and associated duration of symptoms can be predictable (e.g. side effects of chemotherapy treatment) or unpredictable (e.g. relapse in multiple sclerosis). Such recurrent fluctuations in health states can have an important impact on a person’s health-related quality of life. When symptoms vary by time of day, day of the week, or during the month, it is challenging to obtain reliable health-related quality of life estimates for use in assessing costeffectiveness of interventions. The adequacy of the quality of life estimate will be impacted by: (1) the standard recall period associated with the chosen measure (e.g. ‘health today’ for EQ-5D, ‘past month’ for SF-6D) and the way that respondents understand and make judgements about these recall periods, (2) the chosen timepoints for assessing health-related quality of life in relation to the fluctuations in health, and (3) the assumptions used to interpolate between measurement time points and thus calculate the QALYs.

These issues have not received sufficient methodological attention and instead remain poorly accounted for in economic analyses. There is potential for these issues to considerably distort treatment decisions away from the optimal allocation. This paper brings together evidence from health economics,
psychology and behavioural economics to explore these challenges in depth; presents the solutions that have been applied to date; and details a methodological research agenda for measuring QALYs in recurrent fluctuating health states.