Self-reported health status in the general population over 2 decades: variation in EQ-5D-3L in Health Survey for...

Population reference data based on self-reported EQ-5D constitute a valuable resource in a broad range of settings. Within the UK, EQ-5D has been included in national population surveys for over 30 years, notably as part of the Health Survey for England, however, the extent to which such data varies over time remains largely unknown.

Between 1996 and 2014, the Health Survey for England included the original 3-level version of EQ-5D (EQ-5D-3L) and provides data from more than 100,000 respondents; these data form the basis of the present study. Age-stratified rates of reporting any problem in EQ-5D dimensions were computed for men and women; these were standardised using 1996 as the index year. Logit regression was used to examine the influence of year respondent socio-demographic characteristics, and current health status on rates of reporting any problems in EQ-5D-3L. The data of self-rated health status recorded on a 0-100 scale (EQ VAS) was also analyzed.

More than 30% report a problem with Pain/Discomfort with around 20% reporting a problem with Anxiety/Depression, Mobility having a similar problem rate, about 18%. Some 5% of respondents report a problem with Self Care. After an initial fall from 1996, self-reported health remains relatively stable across 10 years, however between 2008 and 2012 rates of reporting any problem increased, particularly in the Anxiety/Depression dimension and amongst older women. Logit regression analysis demonstrates that most of the covariates had statistically significant coefficients, such as age, gender, education, economic activity, income, and long-standing illness/condition.

The study demonstrates the stability of EQ-5D responses over time in HSE data from 1996 to 2014. However, there is evidence of periodic deterioration in health status notably in the years immediately after 2007. Further investigation of this effect could have implications for the interpretation and use of population data based on EQ-5D. The study demonstrates the importance in national surveys of the general population of regular collection of health status data using a standardised measure of health-related quality of life.